Best Models
Command A+ is currently the highest-ranked model for Time to First Token in the available data.
Use the top-ranked model as a starting point, not an automatic decision.
Compare the top choices on price, speed, availability, and your own constraints.
Open the linked result page and methodology before quoting the ranking.
Median time to first token. Lower means the model starts responding sooner, even if total generation speed differs.
| Rank | Model | Provider | Value |
|---|---|---|---|
| #1 | Command A+ | Cohere | 0.16 s |
| #2 | North Mini Code | Cohere | 0.18 s |
| #3 | NVIDIA Nemotron Nano 12B v2 VL (Reasoning) | NVIDIA | 0.22 s |
| #4 | Tiny Aya Global | Cohere | 0.22 s |
| #5 | Llama Nemotron Super 49B v1.5 (Reasoning) | NVIDIA |
This answer uses the exact same metric and live database ranking as the linked Easy Benchmarks leaderboard.
Source: Artificial Analysis Time to First Token. Updated: Jul 16, 2026.
| 0.26 s |
| #6 | Llama Nemotron Super 49B v1.5 (Non-reasoning) | NVIDIA | 0.26 s |
| #7 | Gemini 2.5 Flash-Lite (Non-reasoning) | 0.32 s |
| #8 | Phi-4 Mini Instruct | Microsoft | 0.32 s |
| #9 | Hermes 3 - Llama-3.1 70B | Nous Research | 0.34 s |
| #10 | Command A | Cohere | 0.34 s |
| #11 | Cogito v2.1 (Reasoning) | Deep Cogito | 0.35 s |
| #12 | Phi-4 Multimodal Instruct | Microsoft | 0.37 s |
| #13 | Ministral 3 3B | Mistral | 0.37 s |
| #14 | Gemma 3n E4B Instruct | 0.38 s |
| #15 | Grok Build 0.1 0616 | SpaceXAI | 0.4 s |
| #16 | Qwen3.5 4B (Reasoning) | Alibaba | 0.4 s |
| #17 | Mistral Small 3.2 | Mistral | 0.4 s |
| #18 | NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning) | NVIDIA | 0.41 s |
| #19 | gpt-oss-20b (high) | OpenAI | 0.41 s |
| #20 | Qwen3.5 4B (Non-reasoning) | Alibaba | 0.43 s |
