What’s currentRob’s list

Checked

The newest model anywhere is Glimpse. Elorian released it 3 days ago.

I follow 65 labs and keep only the latest release from each. 11 of them shipped something this week. The fresher the release, the darker the ink.

  1. ElorianGlimpseVisionReasoning

    Glimpse is a multimodal visual reasoning model from Elorian, built for visual understanding and complex visual reasoning in the physical world.

    Released 17 Oct 2026. 262K token context window. $0.60 in and $3.00 out per million tokens. See it on OpenRouter

  2. DeepSeekDeepSeek V4.1 Flash 1017VisionReasoning

    DeepSeek V4.1 Flash 1017 is the October 17, 2026 snapshot of DeepSeek V4.1 Flash, served by DeepSeek under the `deepseek-flash` API model name.

    Released 17 Oct 2026. 1M token context window. $0.15 in and $0.60 out per million tokens. Scores 39.5 on the Artificial Analysis intelligence index. See it on OpenRouter

    Also current

  3. Reflection AIBeam 1.0 501B A23BTextReasoning

    Beam 1.0 is Reflection's open-weight reasoning model for coding and long-horizon agentic workflows.

    Released 16 Oct 2026. 1M token context window. $0.30 in and $1.00 out per million tokens. See it on OpenRouter

  4. InceptionMercury 3TextReasoning

    Mercury 3 is a diffusion-based language model from Inception, and the lab's most capable model to date.

    Released 16 Oct 2026. 1M token context window. $0.10 in and $0.35 out per million tokens. See it on OpenRouter

    Also current

  5. MoondreamMoondream 3.1VisionReasoning

    Moondream 3.1 is a vision-language model from Moondream, the successor to Moondream 3 Preview, with 9B total parameters and 2B active (mixture-of-experts).

    Released 15 Oct 2026. 32K token context window. $0.30 in and $2.50 out per million tokens. See it on OpenRouter

    Also current

  6. MetaMuse Spark OpenVisionReasoning

    Muse Spark Open is a coding-optimized open-weight model.

    Released 15 Oct 2026. 1M token context window. $1.25 in and $4.25 out per million tokens. See it on OpenRouter

    Also current

  7. NVIDIANemotron 3.5 Super VLVisionReasoning

    NVIDIA Nemotron 3.5 Super VL is an open multimodal mixture-of-experts model from NVIDIA, with 12B active parameters out of 120B total.

    Released 15 Oct 2026. 1M token context window. $0.15 in and $0.50 out per million tokens. Scores 26.5 on the Artificial Analysis intelligence index. See it on OpenRouter

    Also current

  8. TencentHy-Memory-30B-A3BText

    Released 15 Oct 2026. See it on Hugging Face

    Also current

  9. Black Forest Labsflux-3-image-klein-pupVision

    Released 14 Oct 2026. See it on Hugging Face

    Also current

  10. Ant GroupMing-3.0-asr-tinyTranscription

    Released 14 Oct 2026. See it on Hugging Face

    Also current

  11. GoogleGemini 4 ArgonVisionReasoning

    Gemini 4 Argon is Google's flagship reasoning model, built for complex software engineering, deep mathematical and scientific problem solving, and autonomous multi-step agents.

    Released 13 Oct 2026. 1M token context window. $2.00 in and $10 out per million tokens. Scores 52.6 on the Artificial Analysis intelligence index. See it on OpenRouter

    Also current

  12. Liquid AILFM2.5-24B-A2BTextReasoning

    LFM2.5-24B-A2B is a mixture-of-experts reasoning model from Liquid AI with 24B total and 2.45B active parameters. It always reasons before answering and is built for agents that call tools.

    Released 12 Oct 2026. 128K token context window. $0.05 in and $0.70 out per million tokens. See it on OpenRouter

    Also current

  13. Alibaba QwenQwen-Image-2.1-TurboImages

    Released 9 Oct 2026. See it on Hugging Face

    Also current

  14. StepFunStep 5 PreviewVisionReasoning

    Step 5 Preview is StepFun's flagship model for agentic work, built on a sparse Mixture-of-Experts architecture (27B active / 600B total parameters).

    Released 8 Oct 2026. 1M token context window. $1.00 in and $2.70 out per million tokens. Scores 43.7 on the Artificial Analysis intelligence index. See it on OpenRouter

  15. MistralVoxtral-Mini-4B-Realtime-ArabicTranscription

    Released 8 Oct 2026. See it on Hugging Face

    Also current

  16. AnthropicClaude Haiku 5.5VisionReasoning

    Claude Haiku 5.5 is Anthropic's small, fast model for high-volume, cost-sensitive work such as summarization, subagents, and browser use.

    Released 7 Oct 2026. 1M token context window. $0.10 in and $0.50 out per million tokens. Scores 43.4 on the Artificial Analysis intelligence index. See it on OpenRouter

    Also current

  17. OpenBMBMiniCPM-V-4.7-35B-A3BVision

    Released 7 Oct 2026. See it on Hugging Face

    Also current

  18. OpenAIGPT-6.1 Sol ProVisionReasoning

    GPT-6.1 Sol Pro is the same underlying model as [GPT-6.1 Sol](https://openrouter.ai/openai/gpt-6.1-sol), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks.

    Released 29 Sept 2026. 1.1M token context window. $2.00 in and $10 out per million tokens. See it on OpenRouter

    Also current

  19. XiaomiMiMo-V2.6-Flash-MOPDText

    Released 27 Sept 2026. See it on Hugging Face

    Also current

  20. Shanghai AI LabIntern-Decision-4BVision

    Released 26 Sept 2026. See it on Hugging Face

    Also current

  21. PerceptronPerceptron Mk1.5VisionReasoning

    Perceptron Mk1.5 is Perceptron's embodied reasoning model for physical agents.

    Released 25 Sept 2026. 37K token context window. $0.15 in and $1.50 out per million tokens. See it on OpenRouter

  22. FireworksEmber-1VisionReasoning

    Ember-1 is a specialized reasoning model from Fireworks Research, built on [Kimi K3](https://openrouter.ai/moonshotai/kimi-k3).

    Released 24 Sept 2026. 1M token context window. $3.00 in and $15 out per million tokens. See it on OpenRouter

  23. Z.aiGLM 5.3 PrimeTextReasoning

    GLM-5.3-Prime is the high-speed variant of Z.ai's GLM-5.3, inheriting its full capabilities while delivering 1.5–2× the output throughput through inference acceleration.

    Released 23 Sept 2026. 1M token context window. $2.80 in and $8.80 out per million tokens. See it on OpenRouter

    Also current

  24. AionLabsAion 3.5 MiniTextReasoning

    Aion 3.5 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models.

    Released 23 Sept 2026. 262K token context window. $0.70 in and $1.40 out per million tokens. See it on OpenRouter

    Also current

  25. UpstageSolar Mini 4TextReasoning

    Solar Mini 4 is Upstage's compact, cost-efficient language model, a 35B-parameter mixture-of-experts with 3B active parameters and a 524K context window.

    Released 23 Sept 2026. 524K token context window. $0.05 in and $0.20 out per million tokens. Scores 24.1 on the Artificial Analysis intelligence index. See it on OpenRouter

    Also current

  26. AppleLensVLM-9BVision

    Released 21 Sept 2026. See it on Hugging Face

  27. xAIGrok 4.7VisionReasoning

    Grok 4.7 is SpaceXAI's flagship model for coding, agentic tasks, and knowledge work, succeeding Grok 4.6.

    Released 21 Sept 2026. 500K token context window. $2.00 in and $6.00 out per million tokens. Scores 46.4 on the Artificial Analysis intelligence index. See it on OpenRouter

    Also current

  28. PrismMLTernary Bonsai 2 27BVisionReasoning

    Bonsai 2 27B is a 27B-parameter reasoning model from PrismML derived from Qwen3.8-27B. It supports coding, mathematics, tool calling, and image understanding with a 262K-token context window.

    Released 18 Sept 2026. 262K token context window. $0.07 in and $0.50 out per million tokens. See it on OpenRouter

  29. Inference.netSchematron V2 TurboText

    Schematron V2 Turbo is a 3B-parameter HTML-to-JSON extraction model from Inference.net. It prioritizes throughput for high-volume extraction workloads.

    Released 12 Sept 2026. 128K token context window. $0.03 in and $0.15 out per million tokens. See it on OpenRouter

    Also current

  30. Sakana AIFugu Ultra v2VisionReasoning

    Fugu Ultra v2 is the higher-performance model in Sakana AI's Fugu family.

    Released 11 Sept 2026. 1M token context window. $5.00 in and $30 out per million tokens. See it on OpenRouter

    Also current

  31. Nex AGINex-N2.5-MiniVisionReasoning

    Nex-N2.5 is an agentic model built to turn goals into working, verified outcomes.

    Released 8 Sept 2026. 262K token context window. $0.02 in and $0.10 out per million tokens. See it on OpenRouter

    Also current

  32. MicrosoftVibeVoice-ASR-Streaming-7BTranscription

    Released 2 Sept 2026. See it on Hugging Face

    Also current

  33. Coheretiny-aya-l2-thinkerText

    Released 2 Sept 2026. See it on Hugging Face

    Also current

  34. ByteDance SeedSeed 2.1 TurboVisionReasoning

    Seed 2.1 Turbo is a multimodal model from ByteDance Seed for coding and long-horizon agent workflows.

    Released 12 Aug 2026. 262K token context window. $0.50 in and $2.50 out per million tokens. See it on OpenRouter

    Also current

  35. IBMgranite-4.2-30bText

    Released 7 Aug 2026. See it on Hugging Face

    Also current

  36. MiniMaxMiniMax-Music3Audio

    Released 7 Aug 2026. See it on Hugging Face

    Also current

  37. MotifMotif-3Text

    Released 7 Aug 2026. See it on Hugging Face

    Also current

  38. MeituanLongCat-Flash-Lite-SparseText

    Released 31 Jul 2026. See it on Hugging Face

    Also current

  39. Thinking MachinesInkling SmallVisionReasoning

    Inkling Small is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 12B active parameters out of 276B total.

    Released 30 Jul 2026. 524K token context window. $0.45 in and $1.20 out per million tokens. Scores 25.7 on the Artificial Analysis intelligence index. See it on OpenRouter

    Also current

  40. LG AI ResearchK-EXAONE-2.0-750B-A37BText

    Released 29 Jul 2026. See it on Hugging Face

    Also current

  41. Swiss AIApertus-v1.5-70BVision

    Released 24 Jul 2026. See it on Hugging Face

    Also current

  42. LightricksLTX-2.5Video

    Released 23 Jul 2026. See it on Hugging Face

  43. PoolsideLaguna S 2.1TextReasoning

    Laguna S 2.1 is the latest coding agent model from [Poolside](<https://poolside.ai/>).

    Released 21 Jul 2026. 1M token context window. $0.09 in and $0.18 out per million tokens. See it on OpenRouter

    Also current

  44. BaiduUnlimited-OCRVision

    Released 19 Jun 2026. See it on Hugging Face

    Also current

  45. Moonshot AIKimi K3VisionReasoning

    Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI.

    Released 13 Jun 2026. 1M token context window. $0.24 in and $12 out per million tokens. Scores 43.6 on the Artificial Analysis intelligence index. See it on OpenRouter

    Also current

  46. ZyphraZONOS2Speech

    Released 11 Jun 2026. See it on Hugging Face

    Also current

  47. Stability AIstable-audio-3-small-musicAudio

    Released 17 May 2026. See it on Hugging Face

    Also current

  48. Hugging Facenanowhale-100mText

    Released 24 Apr 2026. See it on Hugging Face

  49. Essential AIrnj-1.5-instructText

    Released 22 Apr 2026. See it on Hugging Face

  50. ArceeTrinity Large ThinkingTextReasoning

    Trinity Large Thinking is a powerful open source reasoning model from the team at Arcee AI. It shows strong performance in PinchBench, agentic workloads, and reasoning tasks.

    Released 1 Apr 2026. 262K token context window. $0.25 in and $0.80 out per million tokens. Scores 10.8 on the Artificial Analysis intelligence index. See it on OpenRouter

    Also current

  51. RekaReka EdgeVision

    Reka Edge is an extremely efficient 7B multimodal vision-language model that accepts image/video+text inputs and generates text outputs.

    Released 20 Mar 2026. 16K token context window. $0.10 in and $0.10 out per million tokens. See it on OpenRouter

    Also current

  52. Sarvamsarvam-105bText

    Released 3 Mar 2026. See it on Hugging Face

    Also current

  53. WriterPalmyra X5Text

    Palmyra X5 is Writer's most advanced model, purpose-built for building and scaling AI agents across the enterprise.

    Released 21 Jan 2026. 1M token context window. $0.60 in and $6.00 out per million tokens. See it on OpenRouter

  54. Prime IntellectINTELLECT-3.1Text

    Released 20 Jan 2026. See it on Hugging Face

    Also current

  55. Kakaokanana-2-30b-a3b-thinking-2601Text

    Released 14 Jan 2026. See it on Hugging Face

    Also current

  56. TII FalconFalcon-H1-Tiny-R-0.6BText

    Released 12 Jan 2026. See it on Hugging Face

    Also current

  57. AI21AI21-Jamba2-3BText

    Released 6 Jan 2026. See it on Hugging Face

    Also current

  58. NaverHyperCLOVAX-SEED-Omni-8BText

    Released 23 Dec 2025. See it on Hugging Face

    Also current

  59. Nous ResearchNousCoder-14BText

    Released 23 Dec 2025. See it on Hugging Face

    Also current

  60. RelaceRelace SearchText

    The relace-search model uses 4-12 `view_file` and `grep` tools in parallel to explore a codebase and return relevant files to the user request.

    Released 8 Dec 2025. 256K token context window. $1.00 in and $3.00 out per million tokens. See it on OpenRouter

    Also current

  61. AmazonNova 2 LiteVisionReasoning

    Nova 2 Lite is a fast, cost-effective reasoning model for everyday workloads that can process text, images, and videos to generate text.

    Released 2 Dec 2025. 1M token context window. $0.30 in and $2.50 out per million tokens. Scores 13.4 on the Artificial Analysis intelligence index. See it on OpenRouter

    Also current

  62. ServiceNowApriel-1.6-15b-ThinkerVision

    Released 28 Nov 2025. See it on Hugging Face

    Also current

  63. PerplexitybrowsesafeText

    Released 14 Nov 2025. See it on Hugging Face

    Also current

  64. MorphMorph V3 LargeText

    Morph's high-accuracy apply model for complex code edits. ~4,500 tokens/sec with 98% accuracy for precise code transformations.

    Released 7 Jul 2025. 262K token context window. $0.90 in and $1.90 out per million tokens. See it on OpenRouter

    Also current

  65. Snowflakesnowflake-arctic-instructText

    Released 21 Apr 2024. See it on Hugging Face