Models › Low latency
Low latency
For realtime jobs: chat models with a measured first-token time, fastest first.
- Fastest first token114msDeepSeek-V4.1-FlashTogether AI via huggingface.co, 2026-10-26
- Cheapest with tool calls$0.010per 1M inQwen3-Embedding-8BDeepInfra via huggingface.co, 2026-10-26
- Most fine-tuned 2025 base2.6kfine-tunesQwen3-4B-Instruct-2507Hugging Face base-model tags, 2026-10-26
General chat models with a measured first-token time, fastest first. Classifiers and safety models are left out so a guard model can't top a chat list.
The floor
Every view leaves out test fixtures, models whose size didn't parse, and closed models (unless you ask for them). $0 endpoints show as a free tier and never count as cheapest; classifiers and guards never enter a speed sort. How the rank works.
Kind: All kinds · LLMs · Embeddings · Decision models · Closed
89 models match, first 60 shown. Data observed 2026-10-26 07:36 UTC, Hub snapshot 2026-10-26. Hover a figure for its source and date.
| # | Evidence | ||||||
|---|---|---|---|---|---|---|---|
| 1 | DeepSeek-V4.1-Flashdeepseek-ai | 114 ms | 283 | $0.020 | 34 | MIT | |
| 2 | Qwen3.8-27BQwen | 161 ms | 887 | $0.020 | 23 | Apache 2.0 | |
| 3 | Qwen3-30B-A3BQwen | 166 ms | 113 | $0.12 | 1 | Apache 2.0 | |
| 4 | Qwen3-14BQwen | 184 ms | 91 | $0.070 | 4 | Apache 2.0 | |
| 5 | c4ai-command-r7b-12-2024CohereLabs | 201 ms | 106 | — | 1 | CC BY-NC 4.0 | |
| 6 | Qwen3.5-397B-A17BQwen | 204 ms | 90 | $0.39 | 13 | Apache 2.0 | |
| 7 | GLM-5.2zai-org | 205 ms | 210 | $0.032 | 27 | MIT | |
| 8 | Qwen3.5-27BQwen | 207 ms | 85 | $0.20 | 7 | Apache 2.0 | |
| 9 | Muse-Glimmer-30Bmeta-models | 207 ms | 99 | $0.30 | 3 | Apache 2.0 | |
| 10 | gpt-oss-120bopenai | 218 ms | 1095 | $0.030 | 27 | Apache 2.0 | |
| 11 | Inklingthinkingmachines | 227 ms | 152 | $0.95 | 4 | Apache 2.0 | |
| 12 | Qwen3.8-2.4T-A95BQwen | 228 ms | 138 | $2.00 | 7 | Qwen License (qwen3.8-m… | |
| 13 | Hermes-3-Llama-3.1-70BNousResearch | 231 ms | 30 | $0.70 | 2 | Llama 3 Community Licen… | |
| 14 | granite-4.2-8bibm-granite | 234 ms | 35 | $0.060 | 2 | Apache 2.0 | |
| 15 | Qwen3.5-122B-A10BQwen | 259 ms | 122 | $0.26 | 5 | Apache 2.0 | |
| 16 | gemma-3-12b-itgoogle | 262 ms | 65 | $0.040 | 3 | Gemma Terms of Use | |
| 17 | Qwen3-235B-A22BQwen | 265 ms | 13 | $0.20 | 2 | Apache 2.0 | |
| 18 | Qwen3.6-35B-A3BQwen | 276 ms | 176 | $0.050 | 12 | Apache 2.0 | |
| 19 | gpt-oss-20bopenai | 281 ms | 747 | $0.018 | 13 | Apache 2.0 | |
| 20 | Qwen3-4B-Thinking-2507Qwen | 281 ms | 159 | $0.010 | 2 | Apache 2.0 | |
| 21 | Qwen3.5-35B-A3BQwen | 292 ms | 113 | $0.065 | 8 | Apache 2.0 | |
| 22 | Qwen3.6-27BQwen | 298 ms | 77 | $0.20 | 9 | Apache 2.0 | |
| 23 | Qwen3.5-9BQwen | 310 ms | 64 | $0.080 | 8 | Apache 2.0 | |
| 24 | gemma-3-27b-itgoogle | 310 ms | 33 | $0.080 | 6 | Gemma Terms of Use | |
| 25 | Qwen3-32BQwen | 324 ms | 45 | $0.080 | 5 | Apache 2.0 | |
| 26 | MiniMax-M3MiniMaxAI | 327 ms | 117 | $0.22 | 14 | MiniMax Model License (… | |
| 27 | Llama-3.1-8B-Instructmeta-llama | 329 ms | 161 | $0.020 | 4 | Llama 3.1 Community Lic… | |
| 28 | GLM-5.3zai-org | 341 ms | 121 | $0.070 | 38 | GLM License (glm-5.3) | |
| 29 | Qwen2.5-Coder-3B-InstructQwen | 349 ms | 195 | $0.010 | 2 | Qwen License (qwen-rese… | |
| 30 | Qwen3-4B-Instruct-2507Qwen | 353 ms | 147 | $0.010 | 1 | Apache 2.0 | |
| 31 | Qwen3-235B-A22B-Instruct-2507Qwen | 359 ms | 88 | $0.087 | 7 | Apache 2.0 | |
| 32 | Qwen3-8BQwen | 360 ms | 147 | $0.070 | 2 | Apache 2.0 | |
| 33 | GLM-4.7zai-org | 361 ms | 27 | $0.32 | 7 | MIT | |
| 34 | Qwen2.5-Coder-7B-InstructQwen | 365 ms | 145 | $0.010 | 2 | Apache 2.0 | |
| 35 | GLM-5.3-Flashzai-org | 366 ms | 117 | $0.041 | 35 | MIT | |
| 36 | phi-4microsoft | 373 ms | 74 | $0.056 | 2 | MIT | |
| 37 | gemma-4-26B-A4B-itgoogle | 374 ms | 189 | $0.042 | 16 | Apache 2.0 | |
| 38 | DeepSeek-V4-Prodeepseek-ai | 381 ms | 56 | $0.90 | 17 | MIT | |
| 39 | Qwen2.5-Coder-32B-InstructQwen | 385 ms | 47 | $0.060 | 3 | Apache 2.0 | |
| 40 | Qwen2.5-72B-InstructQwen | 385 ms | 37 | $0.36 | 2 | Qwen License (qwen) | |
| 41 | DeepSeek-V4-Flash-Vision-Expdeepseek-ai | 409 ms | 74 | $0.22 | 3 | MIT | |
| 42 | DeepSeek-V4-Pro-0813deepseek-ai | 410 ms | 98 | $0.26 | 22 | MIT | |
| 43 | Ling-3.0-flashinclusionAI | 419 ms | 153 | $0.021 | 2 | MIT | |
| 44 | Llama-4-Scout-17B-16E-Instructmeta-llama | 427 ms | 69 | $0.080 | 4 | Llama Community License… | |
| 45 | GLM-4.6zai-org | 434 ms | 38 | $0.40 | 5 | MIT | |
| 46 | Llama-3.3-70B-Instructmeta-llama | 448 ms | 85 | $0.10 | 12 | Llama 3.3 Community Lic… | |
| 47 | DeepSeek-V3-0324deepseek-ai | 449 ms | 88 | $0.24 | 4 | MIT | |
| 48 | GLM-5.1zai-org | 455 ms | 75 | $0.97 | 15 | MIT | |
| 49 | L3-8B-Stheno-v3.2Sao10K | 457 ms | 117 | $0.050 | 2 | CC BY-NC 4.0 | |
| 50 | Inkling-Smallthinkingmachines | 466 ms | 166 | $0.45 | 2 | Apache 2.0 | |
| 51 | Kimi-K2.5moonshotai | 482 ms | 60 | $0.45 | 7 | Kimi (Modified MIT) (mo… | |
| 52 | Kimi-K2.6moonshotai | 484 ms | 79 | $0.45 | 18 | Kimi (Modified MIT) (mo… | |
| 53 | MiniMax-M2.7MiniMaxAI | 484 ms | 55 | $0.21 | 7 | other | |
| 54 | Kimi-K3moonshotai | 498 ms | 60 | $0.16 | 22 | Kimi K3 (Modified MIT)… | |
| 55 | DeepSeek-R1-Distill-Llama-8Bdeepseek-ai | 503 ms | 150 | $0.050 | 2 | MIT | |
| 56 | Qwen3-Next-80B-A3B-InstructQwen | 537 ms | 77 | $0.072 | 4 | Apache 2.0 | |
| 57 | MiMo-V2.5-ProXiaomiMiMo | 539 ms | 67 | $0.30 | 4 | MIT | |
| 58 | gemma-3-4b-itgoogle | 541 ms | 116 | $0.040 | 3 | Gemma Terms of Use | |
| 59 | DeepSeek-R1-0528deepseek-ai | 546 ms | 44 | $0.50 | 5 | MIT | |
| 60 | Qwen3-VL-30B-A3B-InstructQwen | 556 ms | 40 | $0.15 | 4 | Apache 2.0 |
Evidence marks, left to right: coding, tools, reasoning, long docs, vision, chat. measured on a third-party board, author-reported, structural, none yet. Each mark links to its job from the model page. INT4 is an estimate of the weights alone (method).