Claude Haiku 5.5, Anthropic: 160 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18160
Nemotron 3.5 Lightning, NVIDIA: 147 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18147
Inkling Small, Thinking Machines Lab: 146 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18146
Gemini 3.1 Flash-Lite, Google: 141 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18141
Gemini 3.5 Flash-Lite, Google: 131 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18131
Mistral Medium 3.5, Mistral AI: 129 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18129
DeepSeek V4.1 Flash 1017, DeepSeek: 128 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18128
DeepSeek-V4.1-Flash, DeepSeek: 125 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18125
Gemini 3.5 Flash, Google: 125 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18125
Gemini 3.6 Flash, Google: 122 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18122
MiniMax M2.7, MiniMax: 114 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18114
GPT OSS 120B, OpenAI: 113 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18113
Mistral Small 4, Mistral AI: 109 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18109
GPT-5.4 mini, OpenAI: 108 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18108
Muse Spark 1.3, Meta: 107 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18107
GPT-5 nano, OpenAI: 107 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18107
Claude Sonnet 5.5, Anthropic: 106 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18106
GPT OSS 20B, OpenAI: 105 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-18105
GPT-4.1 nano, OpenAI: 99 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-1899
Grok 4.20 Reasoning, xAI: 95 tokens per second · Median across providers on OpenRouter from the readings of the last 24 hours (each covering the last 30 minutes), excluding time to first token · Retrieved 2026-11-1895