Fastest LLM APIs for Low-Latency Apps
LLM APIs optimized for fast inference and low latency, ranked by measured throughput (tokens/sec). Compare verified pricing on fast inference providers like Groq, Together, and Fireworks.
MODELPRICEWATCH.COM · 2026-11-07
Cost calculator for this use casemonthly cost, top 3 models
- 1 Llama 4 Scout
- $—
- 2 Llama 3.3 70B
- $—
- 3 GPT-OSS 120B
- $—
Full ranking — top 26 models
list prices, USD per 1M tokens| Model | Throughput | Blended* | Input | Output | Context | Provider |
|---|---|---|---|---|---|---|
| 1Llama 4 Scout | 2600 tok/s | $0.150 | $0.100 | $0.300 | 10M | DeepInfra |
| 2Llama 3.3 70B | 2500 tok/s | $1.04 | $1.04 | $1.04 | 131K | Together |
| 3GPT-OSS 120B | 1829 tok/s | $0.262 | $0.150 | $0.600 | 128K | Fireworks |
| 4GPT-OSS 20B | 942 tok/s | $0.128 | $0.070 | $0.300 | 128K | Fireworks |
| 5Gemini 3.5 Flash-Lite | 365 tok/s | $0.850 | $0.300 | $2.50 | 1M | |
| 6GLM-5.2 | 347 tok/s | $2.15 | $1.40 | $4.40 | 1M | Z.AI |
| 7Kimi K2.6 | 343 tok/s | $1.71 | $0.950 | $4.00 | 262K | Moonshot |
| 8Kimi K2.5 | 338 tok/s | $1.20 | $0.600 | $3.00 | 262K | Moonshot |
| 9Nova Micro | 300 tok/s | $0.061 | $0.035 | $0.140 | 128K | Amazon |
| 10Gemini 3.8 Flash | 278 tok/s | $1.50 | $0.750 | $3.75 | 1M | |
| 11Muse Spark 1.3 | 228 tok/s | $2.00 | $1.25 | $4.25 | 1M | Meta |
| 12Muse Spark 1.1 | 206 tok/s | $2.00 | $1.25 | $4.25 | 1M | Meta |
| 13Nova Lite | 200 tok/s | $0.105 | $0.060 | $0.240 | 300K | Amazon |
| 14Codestral 2508 | 200 tok/s | $0.450 | $0.300 | $0.900 | 256K | Mistral |
| 15Gemini 3.6 Flash | 190 tok/s | $1.50 | $0.750 | $3.75 | 1M | |
| 16GPT-5.6 Luna | 185 tok/s | $0.450 | $0.200 | $1.20 | 1M | OpenAI |
| 17GPT-5.4 mini | 180 tok/s | $1.69 | $0.750 | $4.50 | 1M | OpenAI |
| 18DeepSeek V4 Pro | 175 tok/s | $1.63 | $1.30 | $2.60 | 1M | DeepInfra |
| 19Ministral 3 3B | 174 tok/s | $0.100 | $0.100 | $0.100 | 128K | Mistral |
| 20GLM-5 | 170 tok/s | $1.55 | $1.00 | $3.20 | 128K | Z.AI |
| 21Claude Fable 5.1 | 65 tok/s | $20.00 | $10.00 | $50.00 | 1M | Anthropic |
| 22Claude Haiku 5.5 | — | $0.200 | $0.100 | $0.500 | 1M | Anthropic |
| 23MiniMax-M3.1-Flash | — | $0.262 | $0.150 | $0.600 | 1M | MiniMax |
| 24Mercury 3 | — | $0.325 | $0.200 | $0.700 | 1M | Inception |
| 25Mercury 2.5 | — | $0.338 | $0.200 | $0.750 | 260K | Inception |
| 26Kimi K3.1 Pro HighSpeed | — | $8.00 | $4.00 | $20.00 | 1M | Moonshot |
Recent price movement in this ranking
price-only deltas · logged by the daily scan3 of the top 26 fast inference models have re-priced since we began tracking · last checked .
| Model | Changes | Latest move | Provider |
|---|---|---|---|
| Gemini 3.6 Flash | 1 | price cut on : $1.50/$7.50 → $0.750/$3.75 /Mtok | |
| GPT-5.6 Luna | 1 | price cut on : $1.00/$6.00 → $0.200/$1.20 /Mtok | OpenAI |
| Llama 3.3 70B | 1 | price rise on : $0.880/$0.880 → $1.04/$1.04 /Mtok | Together |
Most recently, Gemini 3.6 Flash cut its price on — a sign pricing in this category is still moving, so re-check before committing to a long-term choice. See all recent price moves →
How models are selected
Generally-available models with independently measured throughput (output tokens per second) or a speed-optimized release, ranked by measured throughput. One row per model — host duplicates are collapsed. Models without a speed measurement rank below measured ones, cheapest first.
Prices are per million tokens (Mtok), sourced directly from — and linked to — official provider pricing pages. "Blended cost" is (3×input + 1×output) ÷ 4 — weighted toward input because real workloads read far more tokens than they generate.