Changelog

Changelog

Changelog

1. September 29, 2026Added

   Claude Sonnet 5.5

   Anthropic's Claude Sonnet 5.5 joins the catalog — the direct upgrade to Claude Sonnet 5, at the same $2 / $10 per million input / output tokens, with adaptive thinking on by default, 1M-token context, 128K max output, and vision input.

   Claude Sonnet 5.5Best speed/intelligence mix in the Sonnet tier, successor to Claude Sonnet 5, BYOK-only.
   Release note →
2. September 29, 2026Added

   Step 5 Preview

   StepFun's next-generation flagship base model joins the catalog BYOK-only on AIHubMix — 1M-token context, 64K max output, text/image/video input, and adjustable reasoning at $1.08 / $3.0888 per million input / output tokens.

   Step 5 PreviewStepFun's flagship base model for real-world work, with a programming and professional-expertise focus. BYOK-only via AIHubMix.
   Release note →
3. September 27, 2026Added

   BGE-M3, BGE Multilingual Gemma 2, Mistral Small 3.2 — now on OVHcloud AI Endpoints

   Two BAAI multilingual embedding models and Mistral's current 24B instruction model join the catalog, both on a new BYOK provider: OVHcloud AI Endpoints.

   BGE-M3Multilingual embedding model (100+ languages, 8K context) returning a 1,024-dimension vector.BGE Multilingual Gemma 2Compact Gemma 2 based multilingual embedding model, 8K context, 3,584-dimension vector.Mistral Small 3.2 24B Instruct24B instruction model, 131K context, Apache-2.0.
   Release note →
4. September 26, 2026Added

   Kev 4B

   Kev 4B is a small open-weight decision model — a LoRA adapter and pointer head on Qwen3.5-4B-Base. It answers typed questions about a state with calibrated probabilities instead of generated text, on the same native `/v1/decisions` contract as the rest of the family.

   Kev 4B4B open-weight decision model, 8K context, $0.042 / 1M input and $0 output, BYOK + community pool.
   Release note →
5. September 25, 2026Added

   Decisions family + anyr decision

   Fastino GLiNER2.5-Decide joins TypeSafe Jev and the anyrouter/decision preset, while the new `anyr decision` command gives the whole Decisions family a native terminal client instead of a chat-agent launcher.

   GLiNER2.5-DecideFastino's 340M typed-decision classifier, BYOK + community pool on /v1/decisions. JevTypeSafe's noul / choice / score decision model on the same native contract. AnyRouter DecisionAuto-routed Decisions preset across the systemone models your keys can reach.
   Release note →
6. September 25, 2026Added

   LongCat-2.5-Preview

   Meituan's LongCat-2.5-Preview — a ~1.6T-parameter MoE with ~48B active, 1M-token context, image understanding, and a thinking toggle — joins the catalog over the official LongCat API, BYOK.

   LongCat-2.5-PreviewMeituan's 1.6T/48B MoE preview, 1M context, vision in, BYOK via the official LongCat API.
   Release note →
7. September 24, 2026Added

   Space Bunny Alpha

   Space Bunny Alpha — an anonymous frontier-class stealth preview with blazing-fast inference, strong coding performance, and native multimodal input — joins the catalog as a free model. It always reasons with adjustable effort over a 1M-token context window, served $0 via OpenRouter, Nous, and OpenCode Zen platform hops.

   Space Bunny AlphaAnonymous stealth preview, 1M context, adjustable reasoning, vision input, free on AnyRouter.
   Release note →
8. September 24, 2026Added

   Verified AIHubMix BYOK models

   Five verified text-output models join the catalog through AIHubMix and other BYOK routes: Qwen3 235B A22B Instruct 2507, Mercury 2.5, L3.1 Euryale 70B, GPT-6 Sol, and GPT-6 Luna. Published provider prices and context windows are recorded per upstream; AnyRouter does not add platform markup to BYOK traffic.

   Qwen3 235B A22B (Instruct 2507)Qwen's updated non-thinking MoE model, 262K context, AIHubMix and OpenRouter BYOK.Mercury 2.5Inception's diffusion language model, 260K context, AIHubMix and OpenRouter BYOK.L3.1 Euryale 70BSao10k's creative-roleplay text model, 131K context, AIHubMix and OpenRouter BYOK.GPT-6 SolOpenAI's high-end GPT-6 coding and agent model, 1.05M context, BYOK routes.GPT-6 LunaOpenAI's efficient GPT-6 model for focused, high-throughput tasks, 1.05M context, BYOK routes.
   Release note →
9. September 23, 2026Disabled

   Union Alpha disabled

   Previously verified platform routes return 404 for stealth/union-alpha. Hidden from the catalog; configured BYOK fallbacks remain available for users who saved keys.

   Union AlphaPreviously verified platform routes return 404 for stealth/union-alpha. Hidden from the catalog; configured BYOK fallbacks remain available for users who saved keys.
   Release note →
10. September 22, 2026Added

    Claude Opus 5.5

    Anthropic's Claude Opus 5.5 — the successor to Claude Opus 5 for long-running agentic coding and knowledge work — joins the catalog. It delivers Fable 5.1-level performance at 40% lower cost ($4 / $20 per million input / output tokens), with adaptive thinking always on, 1M-token context, vision input, and tool use.

    Claude Opus 5.5Flagship reasoning/coding model, successor to Claude Opus 5, 40% lower cost, BYOK-only.
    Release note →
11. September 21, 2026Added

    Grok 4.7

    xAI's Grok 4.7 joins the catalog — 500K context, adjustable reasoning, vision in, and six BYOK upstream routes.

    Grok 4.7
    Release note →
12. September 17, 2026Disabled

    Llama Nemotron Embed VL 1B v2 disabled

    NVIDIA's hosted NIM API returns 404 for this catalog id. The model is downloadable self-host NIM only. The live NVIDIA embedding SKU is nvidia/nemotron-3-embed-1b.

    Llama Nemotron Embed VL 1B v2NVIDIA's hosted NIM API returns 404 for this catalog id. The model is downloadable self-host NIM only. The live NVIDIA embedding SKU is nvidia/nemotron-3-embed-1b.
    Release note →
13. September 16, 2026Added

    Union Alpha now on AnyRouter

    Stealth preview Union Alpha joins the catalog free — multimodal research, coding, and agentic workflows at $0.

    Union Alpha
    Release note →
14. September 15, 2026Added

    Agnes 3.0 Flash and Nemotron Parse 2.0

    Agnes AI's Agnes 3.0 Flash (512K, vision in, text out) and NVIDIA Nemotron Parse 2.0 (document images in, structured markdown out) join the catalog.

    Agnes 3.0 FlashAIHubMix BYOK-style $0 billed; list $0.03 in / $0.15 out per 1MNemotron Parse 2.0Vision-in / text-out document parse; not image generation
    Release note →
15. September 15, 2026Added

    AIHubMix BYOK hops on DeepSeek V4.1 Flash, Solar Pro 4, and GPT-6 Astra

    Existing listings pick up an aihubmix-byok hop. Same catalog ids — not new SKUs. Platform aihubmix stays off (#3031).

    DeepSeek V4.1 Flashaihubmix-byok list $0.155 / $0.62 per 1M; billed $0Solar Pro 4aihubmix-byok list $0.3 / $1.2 per 1M; billed $0GPT-6 Astraaihubmix-byok list $10 / $50 per 1M; billed $0
    Release note →
16. September 15, 2026Added

    Context-window floors on anyrouter/auto, free, byok, and the other virtual ids

    Append [1m] or [500k] to any first-party virtual id to keep only members with at least that context window. Not a new listing — the same preset, with a token floor.

    AnyRouter Autoanyrouter/auto[1m] / [500k] AnyRouter Freeanyrouter/free[1m] / [500k] AnyRouter BYOKanyrouter/byok[1m] / [500k] AnyRouter Coding AnyRouter Agent AnyRouter Hermes AnyRouter Cowork AnyRouter Latest
    Release note →
17. September 15, 2026Upstream

    Poolside platform route disabled

    The Poolside platform key returns HTTP 429 usage limit exceeded. Listings stay: poolside/laguna-s-2.1, poolside/laguna-xs-2.1, poolside/laguna-m.1, poolside/laguna-xs.2 — not new SKUs. Platform poolside is off; hops parked. poolside-byok stays live.

    Poolside (platform)HTTP 429 usage limit exceeded (#3297, #3159). Hops parked (#3298). poolside-byok stays on.
    Release note →
18. September 13, 2026Added

    Ling-3.0-flash-VL, Sante, Fin, and OpenRouter :free hops

    inclusionAI Ling-3.0-flash-VL, Sante, and Fin join the catalog, plus Nex-N2.5 Mini/Pro. OpenRouter :free is the wire and an alias — Nemotron 3 Ultra and Inkling Small pick up hue hops so we can burn remaining free quota.

    Ling-3.0-flash-VLVision SKU; OpenRouter :free is an alias, not the listing idLing-3.0-flash-SanteHealth/medicine SKU; Command Code and OpenRouter :free are aliases, not the listing idLing-3.0-flash-FinFinance SKU; OpenRouter :free is an alias, not the listing idNex-N2.5-MiniOpenRouter :free-only SKU; distinct from retired nex-n2-proNex-N2.5-ProOpenRouter :free-only SKU; distinct from retired nex-n2-pro
    Release note →
19. September 10, 2026Added

    DeepSeek V4.1 Flash

    DeepSeek V4.1 Flash joins the catalog as deepseek/deepseek-v4.1-flash, with official DeepSeek API and OpenRouter BYOK hops. Peak list $0.30 / $1.20 per 1M tokens.

    DeepSeek V4.1 FlashOpenRouter listing id; Direct API wire is deepseek-flash
    Release note →
20. September 7, 2026Disabled

    2 models disabled

    Caugiay backend disabled 2026-09-07 (#3146/#2871) — HTTP 401; this model's only platform upstream was caugiay. Plus 1 more.

    Gemma 3 27B ITCaugiay backend disabled 2026-09-07 (#3146/#2871) — HTTP 401; this model's only platform upstream was caugiay.Qwen3.6-27BCaugiay backend disabled 2026-09-07 (#3146/#2871) — HTTP 401; this model's only platform upstream was caugiay.
    Release note →
21. September 7, 2026Upstream

    Caugiay and SambaNova platform routes disabled

    The Caugiay platform key returns HTTP 401 Access denied, and the SambaNova platform account requires a payment method (HTTP 402). BYOK SambaNova stays live; Caugiay has no BYOK sibling. NVIDIA NIM hops that 404 for models NVIDIA does not serve were dropped per listing, not the whole nvidia backend.

    Caugiay (platform)FPT Cloud 401 Invalid API Key (#3146, #2871). SambaNova (platform)HTTP 402 PAYMENT_METHOD_REQUIRED on google/gemma-4-31b (#3172). sambanova-byok stays on.
    Release note →
22. September 5, 2026Added

    Claude Fable 5.1 and GPT-6 Astra on Experiential

    Experiential Cloud promotional hops for Claude Fable 5.1 and a new GPT-6 Astra listing (BYOK live; hosted platform hop pending #2842 recert).

    GPT-6 AstraNew listing — Experiential Cloud promo slug gpt-6-astraClaude Fable 5.1Adds Experiential Cloud promo hop claude-fable-5.1
    Release note →
23. September 5, 2026Disabled

    Pareto disabled

    Cloudflare AI Gateway Unified Billing is exhausted (#3121), so the platform cloudflare backend is disabled. Keep the verified unbiased/pareto model as a disabled catalog tombstone until the backend is funded and re-enabled.

    ParetoCloudflare AI Gateway Unified Billing is exhausted (#3121), so the platform cloudflare backend is disabled. Keep the verified unbiased/pareto model as a disabled catalog tombstone until the backend is funded and re-enabled.
    Release note →
24. September 4, 2026Added

    Muse Spark 1.3

    Meta's Muse Spark 1.3 joins the catalog over AIHubMix BYOK — a multimodal reasoning model for agentic tasks with a 1M-token context window.

    Muse Spark 1.3
    Release note →
25. September 3, 2026Added

    Gemini 3.8 Flash and Mercury 2.5 Preview

    Google Gemini 3.8 Flash and Inception Labs Mercury 2.5 Preview join the catalog. Claude Fable 5.1 and Qwen3.8 Max pick up AIHubMix hops on their existing ids.

    Gemini 3.8 FlashMercury 2.5 Preview
    Release note →
26. September 3, 2026Disabled

    Ox Alpha disabled

    This model has been removed. Requests that still use stealth/ox-alpha or stealth/ox-alpha[1m] are routed to anyrouter/auto.

    Ox AlphaThis model has been removed. Requests that still use stealth/ox-alpha or stealth/ox-alpha[1m] are routed to anyrouter/auto.
    Release note →
27. September 2, 2026Added

    Claude Fable 5.1

    Anthropic Claude Fable 5.1 joins the catalog as one listing, routed through Anthropic BYOK and Cloudflare Workers AI, with OpenRouter BYOK as a fallback.

    Claude Fable 5.1
    Release note →
28. September 2, 2026Added

    Hy4 Preview on AIHubMix BYOK

    Tencent Hunyuan Hy4 Preview joins the catalog as tencent/hy4-preview, routed through AIHubMix BYOK at the aggregator's published list rates.

    Hy4 Preview
    Release note →
29. September 2, 2026Added

    Qwen3.8 Flash on AnyRouter

    Alibaba Qwen3.8 Flash via AIHubMix BYOK. List $0.1126 / $0.38 per 1M.

    Qwen3.8 Flash
    Release note →
30. September 1, 2026Disabled

    2 models disabled

    Every remaining platform host returns 404 for this catalog id. NVIDIA's current Nano omni/VL successor is nvidia/nemotron-3-nano-omni-30b-a3b-reasoning. Plus 1 more.

    Nemotron Nano 12B V2 VLEvery remaining platform host returns 404 for this catalog id. NVIDIA's current Nano omni/VL successor is nvidia/nemotron-3-nano-omni-30b-a3b-reasoning.Nemotron Nano 9B V2Every remaining platform host returns 404 for this catalog id. NVIDIA's current Nano line is listed as nvidia/nemotron-3-nano-30b-a3b.
    Release note →
31. August 28, 2026Disabled

    Hy3 disabled

    Hosted probes for this catalog id fail with a bad request. The model is unlisted rather than advertised as a live hosted chat model.

    Hy3Hosted probes for this catalog id fail with a bad request. The model is unlisted rather than advertised as a live hosted chat model.
    Release note →
32. August 27, 2026Disabled

    5 models disabled

    The remaining platform host is out of credit, and the other platform route is already disabled. Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model. Plus 4 more.

    DeepSeek V3.2The remaining platform host is out of credit, and the other platform route is already disabled. Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model.Qwen3 30B A3BThe platform host no longer serves this catalog id (model_unavailable). Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model.Qwen3.8-Flash-NextThe only upstream for this catalog id is currently skipped by routing (community host unstable). The id is unlisted rather than advertised as a live hosted chat model.QwQ 32BThe platform host no longer serves this catalog id. Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model.Kimi K2.6Wafer platform balance is empty (smoke 2026-08-27). Use moonshotai/kimi-k2.6.
    Release note →
33. August 26, 2026Added

    Catalog sweep — GLM-5.3, vision, Agnes, Sol Disc

    Eight models from the post-closeout sweep: Z.AI GLM-5.3 (coding preview), DeepSeek vision + 0731-fast, Microsoft MAI-Thinking-1, OpenAI GPT-5.6 Sol Disc, and Agnes 2.5 Flash/Pro/Alpha via AIHubMix.

    GLM-5.3DeepSeek-V4-Flash-Vision-ExpMAI-Thinking-1DeepSeek-V4-Flash-0731-FastGPT-5.6 Solwas openai/gpt-5.6-sol-discAgnes 2.5 FlashAgnes 2.5 ProAgnes 2.5 Pro Alpha
    Release note →
34. August 26, 2026Added

    GLM-5.3-Flash — Ox Alpha, named

    The Ox Alpha preview is GLM-5.3-Flash: native multimodal, 1M context, MIT-licensed 320B-A18B. Same AnyRouter routes, catalog id z-ai/glm-5.3-flash.

    GLM-5.3-Flashwas stealth/ox-alpha
    Release note →
35. August 26, 2026Added

    Qwen3.8-Flash-Next from Empero

    Free community endpoint for Qwen3.8-Flash-Next — open by default, from Empero research lab. Catalog id qwen/qwen3.8-flash-next.

    Qwen3.8-Flash-NextEmpero community host
    Release note →
36. August 25, 2026Added

    AIHubMix free tier — 15 more $0 routes

    AIHubMix joins the platform free pool: MiniMax M2.7, Kimi K3, Gemini 3.x Flash, Nemotron 3, GPT-OSS 20B, and more, all routed at $0.

    MiniMax M2.7via aihubmixKimi K3coding-kimi-k3 free idGemini 3.7 Flashfree id now platform-routedGemini 3.6 Flashfree id now platform-routedGemini 3.5 Flash Litefree id now platform-routedNemotron 3 Ultra 550Bfree id now platform-routedNemotron 3 Super 120Bfree id now platform-routedNemotron 3.5 Lightning 30Bfree id now platform-routedDots3-Note Previewfree id now platform-routedLing 3.0 Flashfree id now platform-routed LFM 2.5 2.6Bback — live $0 route againNorth Mini Codefree id now platform-routedLaguna XS 2.1free id now platform-routedOx Alphabare preview id on AIHubMix
    Release note →
37. August 24, 2026Disabled

    2 models disabled

    The only platform host (hoian) is balance-quarantined and no longer serves this catalog id reliably. Unlisted until a live hosted path returns. Plus 1 more.

    Leanstral 1.5The only platform host (hoian) is balance-quarantined and no longer serves this catalog id reliably. Unlisted until a live hosted path returns.Qwen3-Coder-30B-A3BThe platform host no longer serves this catalog id (model_unavailable). Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model.
    Release note →
38. August 21, 2026Added

    Ox Alpha

    Ox Alpha joins the catalog free — a stealth reasoning model for coding, agentic work, and long-horizon engineering.

    Ox Alpha
    Release note →
39. August 20, 2026Added

    Qwen 27B dense SKUs and Llama 3.3 70B Instruct

    Qwen3.6-27B, Qwen3.8-27B, Gemma 3 27B IT, and Meta Llama 3.3 70B Instruct join the catalog as paid chat models.

    Qwen3.6-27BQwen3.8-27BGemma 3 27B IT Llama 3.3 70B Instruct
    Release note →
40. August 20, 2026Added

    Qwen3.8-27B and DeepSeek V4 Flash, 50% off

    Half off through 1 Sep. Same ids, same API — just a friendlier bill.

    Qwen3.8-27B50% off list through 1 Sep 2026DeepSeek-V4-Flash50% off list through 1 Sep 2026
    Release note →
41. August 19, 2026Disabled

    12 models disabled

    The platform Workers AI host no longer serves this catalog id (model_unavailable). Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model. Plus 11 more.

    Gemma SEA-LION v4 27B ITThe platform Workers AI host no longer serves this catalog id (model_unavailable). Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model.Qwen3 MaxThe partner-hosted platform path is no longer billable, and remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model.Qwen 3.7 MaxCloudflare has not published per-token pricing for this partner-hosted SKU (rate visible only in the CF dashboard once provisioned); shipping it live on the metered `cloudflare` backend without a real rate would bill $0.Qwen 3.7 PlusCloudflare has not published per-token pricing for this partner-hosted SKU (rate visible only in the CF dashboard once provisioned); shipping it live on the metered `cloudflare` backend without a real rate would bill $0.Qwen 3.8 MaxCloudflare has not published per-token pricing for this partner-hosted SKU (rate visible only in the CF dashboard once provisioned); shipping it live on the metered `cloudflare` backend without a real rate would bill $0.DeepSeek R1 Distill Qwen 32BThe platform Workers AI host no longer serves this catalog id. Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model.DiffusionGemma 26B A4BChat completions return empty content for this discrete-diffusion model under normal chat token budgets. It is unlisted rather than advertised as a live chat model.Gemma 3 12B ITThe platform Workers AI host no longer serves this catalog id. Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model. Llama 3.2 1B InstructThe platform Workers AI host no longer serves this catalog id (model_unavailable). Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model. Llama 3.2 3B InstructThe platform Workers AI host no longer serves this catalog id (model_unavailable). Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model.Qwen3-32BThe platform host no longer serves this catalog id (model_unavailable). Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model.Qwen3 Embedding 0.6BThe platform Workers AI host no longer serves this catalog id (model_unavailable). Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted embedding model.
    Release note →
42. August 18, 2026Disabled

    Mistral 7B Instruct v0.1 disabled

    Cloudflare Workers AI retired this SKU on 2026-05-30; live smoke 2026-08-14/18 returns model_unavailable (no_route). It was this model's only upstream, so it was disabled rather than the route pruned.

    Mistral 7B Instruct v0.1Cloudflare Workers AI retired this SKU on 2026-05-30; live smoke 2026-08-14/18 returns model_unavailable (no_route). It was this model's only upstream, so it was disabled rather than the route pruned.
    Release note →
43. August 18, 2026Upstream

    GitHub Models platform and BYOK routes disabled

    GitHub Models was fully retired on 2026-07-30. The inference API and BYOK endpoints return HTTP 410, so the github and github-byok backends are off.

    GitHub Models (platform)Retired 2026-07-30 — HTTP 410 on every try (#2237).GitHub Models (BYOK)Retired with the platform API on 2026-07-30 (#2237).
    Release note →
44. August 17, 2026Added

    DeepSeek V4 Pro

    DeepSeek V4 Pro (0813) joins the catalog, hosted on Cloudflare Workers AI with a 1M-token context window.

    DeepSeek V4 Prowas deepseek-ai/deepseek-v4-pro-0813
    Release note →
45. August 17, 2026Disabled

    4 models disabled

    Cloudflare has not published per-token pricing for this partner-hosted SKU (rate visible only in the CF dashboard once provisioned); shipping it live on the metered `cloudflare` backend without a real rate would bill $0. Plus 3 more.

    Qwen 3.5 397B A17BCloudflare has not published per-token pricing for this partner-hosted SKU (rate visible only in the CF dashboard once provisioned); shipping it live on the metered `cloudflare` backend without a real rate would bill $0.DeepSeek V4 Flash 0731 (NIM)NVIDIA NIM smoke timeout 2026-08-17; 99% of prod errors are $0-balance with no serving BYOK. Use deepseek/deepseek-v4-flash.Qwen3.5 397B A17BWafer 404 on every prod generation and smoke 502 (2026-08-17). No other non-BYOK host.MiniMax M3Wafer smoke failed repeatedly (2026-08-14). Use minimax/m3.
    Release note →
46. August 14, 2026Added

    Gemini 3.7 Flash and latest frontier chat SKUs

    Google Gemini 3.7 Flash joins the catalog, plus Qwen3.8 2.4T A95B, Muse Spark 1.2, Inkling Small, Solar Pro 4, Seed 2.1 Turbo, Seed 2.0 Code, Claude Opus 5 Fast, and Dots3-Note Preview.

    Gemini 3.7 FlashQwen3.8 2.4T A95B Muse Spark 1.2Inkling SmallSolar Pro 4 Seed 2.1 Turbo Seed 2.0 CodeClaude Opus 5 FastDots3-Note Preview
    Release note →
47. August 14, 2026Disabled

    5 models disabled

    CF Workers AI LoRA host 404 / model_unavailable on every smoke probe (2026-08-14). Canonical Gemma instruct SKUs remain (gemma-3 / gemma-4). Sibling gemma-2b-it-lora already disabled. Plus 4 more.

    Gemma 7B IT LoRACF Workers AI LoRA host 404 / model_unavailable on every smoke probe (2026-08-14). Canonical Gemma instruct SKUs remain (gemma-3 / gemma-4). Sibling gemma-2b-it-lora already disabled. Llama 3.1 405B InstructGitHub Models retired 2026-07-30 (HTTP 410; smoke 2026-08-14). Only platform backend was github. Do not re-enable unless a non-GitHub free/platform route appears. Llama 2 7B Chat HF LoRAWorkers AI LoRA host — sibling Gemma LoRA SKUs already 404 on this account. Disabled so smoke does not pick it as the next Cloudflare cheapest model.Mistral 7B Instruct v0.2 LoRAWorkers AI LoRA host 404 / model_unavailable (smoke 2026-08-14). Same class as gemma-7b-it-lora and llama-2-7b-chat-hf-lora. Big Pickleopencode-zen 502 on every 2026-08-14 smoke probe. Catalog-only connectivity toy; disable so the backend is not counted as a failing smoke target.
    Release note →
48. August 13, 2026Disabled

    7 models disabled

    SiliconFlow no longer lists this id; its catalogue carries gemma-4-12B-it, gemma-4-26B-A4B-it and gemma-4-31B-it instead. Plus 6 more.

    Gemma-4-27B-itSiliconFlow no longer lists this id; its catalogue carries gemma-4-12B-it, gemma-4-26B-A4B-it and gemma-4-31B-it instead.Nemotron 3.5 NanoIts only upstream went offline and no other provider serves this model — the NVIDIA Nemotron 3.5 line is now published as Lightning and Content Safety only. Nemotron 3.5 Lightning is the closest replacement.GPT-5 ChatCloudflare Workers AI retired this SKU; the binding returns HTTP 410 (Gone) on every attempt. In the same 2026-07-25 smoke run, the sibling partner-hosted model openai/gpt-5.6-luna reached the same OpenAI-on-Workers-AI rail and got HTTP 400 (request rejected, not gone), showing the retirement is specific to this SKU rather than a rail-wide outage or billing issue. It was this model's only upstream, so it was disabled rather than the route pruned.GPT-5.1 ChatCloudflare Workers AI retired this SKU; the binding returns HTTP 410 (Gone) on every attempt. In the same 2026-07-25 smoke run, the sibling partner-hosted model openai/gpt-5.6-luna reached the same OpenAI-on-Workers-AI rail and got HTTP 400 (request rejected, not gone), showing the retirement is specific to this SKU rather than a rail-wide outage or billing issue. It was this model's only upstream, so it was disabled rather than the route pruned.Laguna M.1Poolside's live listing now carries only Laguna XS 2.1 and Laguna S 2.1 — this generation has been withdrawn, and every route to it (including the aggregator mirrors) now 404s. Laguna S 2.1 is the direct replacement.Laguna XS.2Poolside's live listing (inference.poolside.ai/v1/models) now carries only laguna-xs-2.1 and laguna-s-2.1 — this generation is gone, so every upstream here 404s.GLM-4.6V-FlashNo upstream currently serves this model — its only provider went offline, and Z-AI's own API publishes the text GLM line only, not the vision variants. GLM-4.6V is the closest available alternative.
    Release note →
49. August 12, 2026Added

    Nemotron Lightning, Muse Glimmer, Seedance 2.5, DeepSeek NIM Flash

    NVIDIA NIM additions (Nemotron 3.5 Lightning 30B, Meta Muse Glimmer 30B, DeepSeek-V4-Flash 0731), ByteDance Seedance 2.5 on Workers AI, and free Requesty route for Ling-3.0-tiny.

    Nemotron 3.5 Lightning 30B A3BSparse MoE chat model on NVIDIA NIM. Muse Glimmer 30BMeta Muse Glimmer via NVIDIA NIM.DeepSeek V4 Flash 0731Dated NIM packaging of DeepSeek-V4-Flash. Seedance 2.5Audio-video generation on Cloudflare Workers AI.
    Release note →
50. August 12, 2026Added

    Free-route expansion + Liquid LFM 2.5

    Expanded free platform routes across NVIDIA Nemotron free models, GPT-OSS-20B free, Ling-3.0-tiny on Novita, and new Liquid LFM 2.5 2.6B free. CommandCode free/promo routes for Laguna S 2.1 and MiMo V2.5 Pro.

    LFM 2.5 2.6BFree tier via platform free pool + BYOK free SKU.Nemotron 3.5 Lightning 30B A3BAdded free aggregator + free-pool routes.GPT-OSS 20BFree aggregator SKU added.Ling-3.0-tinyNovita platform/BYOK free routes.
    Release note →
51. August 12, 2026Added

    Grok 4.6

    xAI Grok 4.6 joins the catalog via xAI BYOK and OpenRouter BYOK. 500K context, $2/$6 list on xAI.

    Grok 4.6
    Release note →
52. August 12, 2026Disabled

    3 models disabled

    The catalogue is text/embedding-output only; video models stay on disk for future re-enable (same posture as seedance-2.0-mini). Unpriced CF video would also fail metered-catalog-billing. Plus 2 more.

    Seedance 2.5The catalogue is text/embedding-output only; video models stay on disk for future re-enable (same posture as seedance-2.0-mini). Unpriced CF video would also fail metered-catalog-billing.EmbeddingGemma 300MSmoke reported model_unavailable (no upstream for platform keys). See #1799/#1806.Gemma 2B IT LoRAPlatform smoke and e2e hit model_unavailable — the Workers AI binding returns no usable upstream for this LoRA host model under the current account neuron/catalog posture. Canonical Gemma instruct SKUs remain available (gemma-3 / gemma-4 family). See #1799/#1806.
    Release note →
53. August 7, 2026Added

    EmbeddingGemma 300M and Ling-3.0-tiny

    Google's EmbeddingGemma 300M open embedding model and inclusionAI's Ling-3.0-tiny free MoE chat model join the catalog.

    EmbeddingGemma 300M300M multilingual text embeddings on Workers AI.Ling-3.0-tiny7.9B MoE (1.3B active) free instruct model via OpenRouter.
    Release note →
54. August 3, 2026Added

    Sakana Namazu

    Sakana AI's Japanese-specialized LLM (built on Kimi K2.6) joins the catalog via BYOK.

    Sakana NamazuJapanese business-focused LLM with web search and code execution.
    Release note →
55. August 1, 2026Added

    Nemotron 3.5 Nano

    NVIDIA's Nemotron 3.5 Nano joins the catalog via Blackbox AI BYOK.

    Nemotron 3.5 NanoCompact, efficient chat model via Blackbox AI BYOK.
    Release note →
56. July 31, 2026Added

    Nemotron Nano, GLM-4.5/4.6, MiniMax M2, Qwen3.5-9B/27B

    Eight new models discovered from upstream providers and added to the catalog — NVIDIA's free Nemotron Nano 12B V2 VL and 9B V2 via OpenRouter, Z-AI's GLM-4.5 and GLM-4.6 (with vision variant), MiniMax M2, and Alibaba's Qwen3.5-9B and Qwen3.5-27B.

    Nemotron Nano 12B V2 VLFree multimodal reasoning model (text+image+video), 128K context.Nemotron Nano 9B V2Free reasoning model, 128K context, knowledge cutoff 2025-03-31.GLM-4.5355B MoE agent model, 128K context, 96K output.GLM-4.6200K context, stronger coding and reasoning than GLM-4.5.GLM-4.6VVision variant of GLM-4.6, 200K context. MiniMax M2Agentic coding and office productivity model.Qwen3.5-9B262K context, vision support, via DeepInfra and BYOK.Qwen3.5-27BDense frontier model, 262K context, reasoning and coding.
    Release note →
57. July 29, 2026Added

    Qwen3.7 Flash

    Alibaba's Qwen3.7 Flash — a vision-language reasoning model for multimodal agents, visual coding, search, and computer interaction — joins the catalog via OpenRouter BYOK.

    Qwen3.7 FlashVision-language reasoning model, 1M context, OpenRouter BYOK.
    Release note →
58. July 26, 2026Disabled

    16 models disabled

    The catalogue was narrowed to models whose output is text or embeddings; this model's output is an image, so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend). Plus 15 more.

    FLUX.2 DevThe catalogue was narrowed to models whose output is text or embeddings; this model's output is an image, so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend). FLUX.2 Klein 9BThe catalogue was narrowed to models whose output is text or embeddings; this model's output is an image, so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend). Seedream 5 ProThe catalogue was narrowed to models whose output is text or embeddings; this model's output is an image, so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend). Eleven Flash v2.5The catalogue was narrowed to models whose output is text or embeddings; this model's output is audio (text-to-speech), so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend). Eleven Multilingual v2The catalogue was narrowed to models whose output is text or embeddings; this model's output is audio (text-to-speech), so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend).Krea 2 LargeThe catalogue was narrowed to models whose output is text or embeddings; this model's output is an image, so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend).Krea 2 Medium TurboThe catalogue was narrowed to models whose output is text or embeddings; this model's output is an image, so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend).Krea 2 MediumThe catalogue was narrowed to models whose output is text or embeddings; this model's output is an image, so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend). Llama 2 7B Chat FP16Cloudflare Workers AI retired this SKU on 2026-05-30; the binding returns HTTP 410 (deprecated) and the model is absent from the account's live catalogue. It was this model's only upstream, so it was disabled rather than the route pruned. Llama 2 7B Chat INT8Cloudflare Workers AI retired this SKU on 2026-05-30; the binding returns HTTP 410 (deprecated) and the model is absent from the account's live catalogue. It was this model's only upstream, so it was disabled rather than the route pruned. Llama 3 8B Instruct AWQCloudflare Workers AI retired this SKU on 2026-05-30; the binding returns HTTP 410 (deprecated) and the model is absent from the account's live catalogue. It was this model's only upstream, so it was disabled rather than the route pruned. Llama 3 8B InstructCloudflare Workers AI retired this SKU on 2026-05-30; the binding returns HTTP 410 (deprecated) and the model is absent from the account's live catalogue. It was this model's only upstream, so it was disabled rather than the route pruned. Llama 3.1 8B Instruct AWQCloudflare Workers AI retired this SKU on 2026-05-30; the binding returns HTTP 410 (deprecated) and the model is absent from the account's live catalogue. It was this model's only upstream, so it was disabled rather than the route pruned. Llama 3.1 8B Instruct FastThe model is gone from Cloudflare's live account catalogue and the binding returns HTTP 404 (model does not exist). It was this model's only upstream, so it was disabled rather than the route pruned.Grok STTThe catalogue was narrowed to text- and embedding-output models. This model's output IS text, but its input is audio — it is a speech-to-text SKU, named explicitly in the same decision, so it goes out of service with the image / video / TTS models rather than being kept on the technicality of its output modality.Grok TTSThe catalogue was narrowed to models whose output is text or embeddings; this model's output is audio (text-to-speech), so it was taken out of service. Related: #1551 (these SKUs bill $0 on a metered backend).
    Release note →
59. July 25, 2026Added

    Claude Opus 5

    Anthropic's Claude Opus 5 — the flagship model for demanding reasoning, coding, and long-horizon agentic work — joins the catalog, BYOK-only like its Claude siblings.

    Claude Opus 5Flagship reasoning/coding model, BYOK-only.
    Release note →
60. July 24, 2026Added

    Ling-3.0-flash

    inclusionAI's Ling-3.0-flash — a 124B-parameter Mixture-of-Experts model (~5.1B active) tuned for token-efficient, production-scale agentic inference — joins the catalog, served free with a user-owned API key (BYOK).

    Ling-3.0-flash124B MoE instant model, free via BYOK.
    Release note →
61. July 24, 2026Upstream

    Google unified-billing and Blackbox platform routes disabled

    The Cloudflare AI Gateway unified-billing balance and the Blackbox platform budget both ran out. BYOK siblings stay live; platform traffic fails over or stops.

    Google AI (platform)Unified billing 402 / insufficient balance (#1321). No recharge.Blackbox AI (platform)Provider budget cap (~$3). No traffic until the account is topped up (#1326).
    Release note →
62. July 22, 2026Added

    Gemini 3.6 Flash & Gemini 3.5 Flash Lite

    Google's Gemini 3.6 Flash and Gemini 3.5 Flash Lite join the catalog as partner-hosted SKUs on Cloudflare Workers AI, with OpenRouter BYOK fallback.

    Gemini 3.6 FlashGemini 3.5 Flash Lite
    Release note →
63. July 21, 2026Added

    Laguna S 2.1

    Poolside's latest coding agent model Laguna S 2.1 (118B total, 8B active parameters, 1M context, 70.2% Terminal-Bench 2.1, 40.4% DeepSWE) joins the catalog — open-weight under OpenMDW-1.1.

    Laguna S 2.1
    Release note →
64. July 17, 2026Added

    Inkling + Nemotron 3 Embed 1B

    Thinking Machines Lab's first open-weights foundation model Inkling (975B MoE, 41B active, 1M context, multimodal input with switchable reasoning) and NVIDIA's Nemotron-3-Embed-1B multilingual embedding model (2048 dims, 34 languages) join the catalog via NVIDIA NIM.

    InklingNemotron 3 Embed 1BMultilingual text embeddings (2048 dimensions, 32k context) for search and RAG.
    Release note →
65. July 16, 2026Added

    Kimi K3

    Moonshot AI's Kimi K3 is now available on AnyRouter — an open-source model with a 1M-token context window and native visual understanding (text, image, and video). Built for long-horizon software engineering and deep reasoning where code meets visual and spatial thinking, with thinking mode always on.

    Kimi K3 Muse Spark 1.1Meta's multimodal agentic reasoning model (text/image/video/audio/PDF in, 1M context) via OpenRouter BYOK. US-only.
    Release note →
66. July 16, 2026Upstream

    OpenAI and DeepInfra platform routes disabled

    The platform OpenAI key has no quota (429 insufficient_quota) and DeepInfra has no payment method (402). BYOK users are unaffected.

    OpenAI (platform)No quota — 5,179 attempts / 0 ok over 30d (#1283). DeepInfra (platform)No payment method — 1,240 attempts, 76.9% fail (#1284).
    Release note →
67. July 10, 2026Upstream

    Named OpenRouter platform backend disabled

    The public `openrouter` backend is off. Platform traffic uses the masked hue route; user keys stay on openrouter-byok.

    OpenRouter (named platform)Disabled #1152. Hue is the platform path; openrouter-byok is BYOK.
    Release note →
68. July 9, 2026Added

    Grok 4.5 and Laguna XS 2.1

    Two frontier additions: xAI's Grok 4.5 joins the catalog and Poolside ships a refreshed Laguna XS 2.1 coding model.

    Grok 4.5Laguna XS 2.1
    Release note →
69. July 8, 2026Added

    Free-tier open models via SiliconFlow

    A batch of open-weight models now has a free-tier upstream — Qwen3 across sizes, Gemma 4 27B, and Tencent Hy3.

    Qwen3-32BQwen3-14BQwen3-8BQwen3-Coder-30B-A3BCoding-tuned MoEGemma 4 27BHy3
    Release note →
70. July 2, 2026Added

    Claude Sonnet 5

    Anthropic's Claude Sonnet 5 is available through BYOK — bring your Anthropic key and route it behind the gateway.

    Claude Sonnet 5BYOK
    Release note →
71. July 1, 2026Added

    Codex-class OpenAI models and GLM-5

    The OpenAI Codex family lands — GPT-5.3 / 5.2 Codex and 5.1 Codex Max — alongside Z.ai's GLM-5 and Qwen3.7 Max.

    GPT-5.3 CodexGPT-5.2 CodexGPT-5.1 Codex MaxGLM-5Qwen3.7 Max
    Release note →
72. June 30, 2026Added

    LongCat-2.0

    The LongCat provider joins the catalog with LongCat-2.0, available via BYOK.

    LongCat-2.0BYOK
    Release note →
73. June 24, 2026Added

    Cerebras, Venice, and Wafer providers

    Three new upstreams broaden coverage — Cerebras GPT-OSS, Venice's MiniMax M2.7, and Wafer's aggregated MiniMax M3 and GLM-5.2.

    MiniMax M2.7MiniMax M3GLM-5.2Kimi K2.7 Code
    Release note →
74. Date not recordedDisabled

    13 models disabled

    Its BYOK aggregator upstream no longer lists this model id (verify:upstreams MISSING 2026-07-16), and that was its only upstream. See #1283. Plus 12 more.

    Seedance 2.0 MiniMAI-Image-2.5Its BYOK aggregator upstream no longer lists this model id (verify:upstreams MISSING 2026-07-16), and that was its only upstream. See #1283.Nex-N2-ProCosmos 3 Nano ReasonerNVIDIA has not exposed this on the hosted API yet (integrate.api.nvidia.com 400s); it ships as self-host NIM + HF weights only.Ising Calibration 1.5 31BInkling 256KStaged, not served (#1512): Cloudflare publishes no per-token price for this SKU, and vendor guidance says it's intended for low-traffic testing/internal use, not high-throughput production.Kimi K2.7 CodeGLM-5.2Qwen3.6 35B A3BQwen3.7 MaxGLM-5.2Grok Imagine Video 1.5 (Preview)Grok Voice TTS 1.0Its BYOK aggregator upstream no longer lists this model id (verify:upstreams MISSING 2026-07-16), and that was its only upstream. See #1283.
    Release note →