Providers, measured.
Every model served by two or more providers, with the provider that leads each figure: price, output speed, time to first token and uptime, measured on llm.buzz. Open a model to compare its providers.
42 models · 103 providers · figures through Oct 27, 04:00 UTC
| Model | Cheapest$ per 1M tokens | Fastest outputtok/s, median over 24 h | Quickest first tokenseconds, median over 24 h | Best uptimeshare served over 72 h |
|---|---|---|---|---|
| Kimi K3 | $5.13 blended llm.buzz | 78 tok/stest DeepInfra | 0.83 stestDeepInfra | 100%testDeepInfra |
| GLM 5.3 | $1.50 blendedllm.buzz | 278 tok/stest Prime Intellect | 0.28 stest Scaleway | 100%testPrime Intellect |
| Qwen3.8 27B | $0.775 blendedDeepInfra | 161 tok/stestScaleway | 0.23 stestScaleway | 100%testScaleway |
| Qwen3.8 Flash | $0.1725 blendedllm.buzz | 78 tok/stestllm.buzz | 1.20 stest Novita | 100%testNovita |
| GLM 5.3 Flash | $0.2375 blendedllm.buzz | 93 tok/stestllm.buzz | 0.69 stestPrime Intellect | 100%testPrime Intellect |
| DeepSeek V4 Flash 0731 | $0.09 blendedDeepInfra | 134 tok/stestScaleway | 0.40 stestScaleway | 100%testScaleway |
| DeepSeek V4.1 Flash | $0.30 blendedDeepInfra | 193 tok/stestPrime Intellect | 0.40 stestScaleway | 100%testPrime Intellect |
| DeepSeek V4.1 Flash 1017 | $0.225 blendedllm.buzz | 193 tok/stestPrime Intellect | 0.48 stestPrime Intellect | 100%testPrime Intellect |
| Gemma 4 26B A4B | $0.1375 blendedDeepInfra | 232 tok/stestScaleway | 0.23 stestScaleway | 100%testDeepInfra |
| gpt-oss-120b | $0.10 blendedllm.buzz | 481 tok/stestllm.buzz | 0.23 stestScaleway | 100%testScaleway |
| Qwen3.6 35B A3B | $0.3125 blendedDeepInfra | 161 tok/stestPrime Intellect | 0.28 stestScaleway | 100%testDeepInfra |
| Claude Fable 5 | $15.00 blendedllm.buzz | 278 tok/stest Anthropic | 4.29 stestAnthropic | 100%testAnthropic |
| Claude Fable 5.1 | $15.00 blendedllm.buzz | 278 tok/stestAnthropic | 2.98 stestAnthropic | 100%testAnthropic |
| Claude Haiku 5.5 | $0.20 blendedllm.buzz | 193 tok/stestllm.buzz | 2.48 stestllm.buzz | 100%testAnthropic |
| Claude Opus 4.7 | $7.50 blendedllm.buzz | 65 tok/stestllm.buzz | 0.69 stestAnthropic | 98.6%testllm.buzz |
| Claude Opus 4.8 | $7.50 blendedllm.buzz | 65 tok/stestllm.buzz | 1.00 stestAnthropic | 100%testAnthropic |
| Claude Opus 5 | $7.50 blendedllm.buzz | 232 tok/stestAnthropic | 3.57 stestAnthropic | 100%testllm.buzz |
| Claude Opus 5.5 | $6.00 blendedllm.buzz | 334 tok/stestAnthropic | 2.98 stestAnthropic | 100%testAnthropic |
| Claude Sonnet 5 | $3.00 blendedllm.buzz | 78 tok/stestllm.buzz | 1.20 stestAnthropic | 100%testAnthropic |
| Claude Sonnet 5.5 | $4.00 blendedllm.buzz | 997 tok/stestAnthropic | 2.98 stestAnthropic | 100%testAnthropic |
| DeepSeek V4 Flash Vision Exp | $0.495 blendedllm.buzz | 193 tok/stestDeepInfra | 0.83 stestDeepInfra | 100%testDeepInfra |
| DeepSeek V4 Pro 0813 | $1.63 blendedllm.buzz | 65 tok/stestllm.buzz | 1.20 stestDeepInfra | 100%testDeepInfra |
| Gemma 4 31B | $0.1875 blendedllm.buzz | 65 tok/stestllm.buzz | 0.69 stestDeepInfra | 100%testDeepInfra |
| Mistral Small 3.2 24B | $0.1729 blendedllm.buzz | 134 tok/stestllm.buzz | 0.19 stestScaleway | 100%testScaleway |
| GPT-4.1 | $3.50 blendedllm.buzz | 334 tok/stestllm.buzz | 1.20 stest OpenAI | 100%testOpenAI |
| GPT-4.1 Mini | $0.63 blendedllm.buzz | 134 tok/stestllm.buzz | 1.00 stestOpenAI | 100%testOpenAI |
| GPT-4o-mini | $0.2625 blendedllm.buzz | 93 tok/stestllm.buzz | 0.83 stestOpenAI | 100%testOpenAI |
| GPT-5 Mini | $0.6875 blendedllm.buzz | 93 tok/stestOpenAI | 2.48 stestOpenAI | 100%testOpenAI |
| GPT-5 Nano | $0.1237 blendedllm.buzz | 161 tok/stestOpenAI | 2.48 stestOpenAI | 100%testOpenAI |
| GPT-5.4 | $5.63 blendedllm.buzz | 65 tok/stestllm.buzz | 1.00 stestOpenAI | 100%testOpenAI |
| GPT-5.4 Mini | $1.69 blendedllm.buzz | 134 tok/stestllm.buzz | 1.00 stestOpenAI | 100%testOpenAI |
| GPT-5.4 Nano | $0.4625 blendedllm.buzz | 193 tok/stestllm.buzz | 0.69 stestOpenAI | 100%testOpenAI |
| GPT-5.5 | $11.25 blendedllm.buzz | 232 tok/stestllm.buzz | 2.07 stestOpenAI | 100%testOpenAI |
| GPT-5.6 Luna | $0.45 blendedllm.buzz | 232 tok/stestllm.buzz | 1.72 stestOpenAI | 100%testOpenAI |
| GPT-5.6 Sol | $8.00 blendedllm.buzz | 78 tok/stestOpenAI | 3.57 stestOpenAI | 100%testOpenAI |
| GPT-5.6 Terra | $4.05 blendedllm.buzz | 93 tok/stestOpenAI | 1.44 stestOpenAI | 98.6%testOpenAI |
| GPT-6 Astra | $20.00 blendedllm.buzz | 31 tok/stestOpenAI | 1.72 stestOpenAI | 100%testOpenAI |
| GPT-6 Luna | $0.20 blendedllm.buzz | 401 tok/stestllm.buzz | 2.48 stestOpenAI | 100%testllm.buzz |
| GPT-6 Sol | $4.00 blendedllm.buzz | 93 tok/stestllm.buzz | 4.29 stestOpenAI | 100%testOpenAI |
| GPT-6.1 Sol | $4.00 blendedllm.buzz | 31 tok/stestllm.buzz | 2.07 stestOpenAI | 100%testOpenAI |
| Surogate 3.7 Flash | $0.2625 blended Simpleminds EU | 112 tok/stestSimpleminds EU | 0.04 stestSimpleminds EU | 100%testSimpleminds EU |
| MiMo-V2.6-Pro | $0.54 blendedllm.buzz | 37 tok/stestllm.buzz | 1.20 stestDeepInfra | 97.1%testllm.buzz |
Figures tagged "test" are from llm.buzz's test requests, where customer traffic is too thin to publish.
How we measure
- Speed and first tokenOutput speed and time to first token are medians over the last 24 hours, per provider, from text requests.
- UptimeThe share of attempts the provider served over the 72 hours before the figures' end, provider failures counted against it.
- Two sourcesCustomer traffic once enough organizations count toward a figure; below that, llm.buzz's test requests, short streamed requests sent to every live provider. The tables tag those "test".
- PriceEach provider's price per million tokens, with a running discount applied. The blended price weighs input and output 3 to 1, a chat request's usual mix.
Let the router pickName the model; llm.buzz sends each request to a provider and moves to the next when one fails.