The Known Good Updated 8 Sep 2026 Subscribe
Providers· Inference host· 27 tracked models

CoreWeave — pricing & performance

Compare models All providers Data sources
Models served
27
of 1,109 tracked models
Mean output speed
88
tokens per second, averaged over
27 measured endpoints
Mean latency
0.5s
time to first token, averaged over
27 measured endpoints
Cheapest blended
$0.055
per 1M tokens at a 3:1
input:output blend
Output speed on CoreWeave

Output tokens per second for all 27 models on CoreWeave with published throughput telemetry, the most recent figure OpenRouter publishes for each host. Higher is better. We do not run these measurements ourselves.

⇩
245
212
154
140
112
104
102
100
99
94
92
84
The Known Good
27 models on CoreWeave
Full leaderboard →
Model Creator Input $/M Output $/M Blended Speed TTFT Context
DeepSeek: DeepSeek V3.1 DeepSeek $0.55 $1.65 $0.82 46 0.7s 161k
DeepSeek: DeepSeek V4 Flash 0423 DeepSeek $0.14 $0.28 $0.17 23 0.5s 1.05M
DeepSeek: DeepSeek V4 Flash 0731 DeepSeek $0.13 $0.28 $0.17 84 0.5s 262k
DeepSeek: DeepSeek V4 Pro 0423 DeepSeek $1.15 $2.55 $1.50 55 0.5s 1.05M
DeepSeek: DeepSeek V4 Pro 0813 DeepSeek $1.31 $3.96 $1.97 99 0.8s 1.05M
Google: Gemma 4 31B Google $0.10 $0.34 $0.16 47 0.6s 262k
IBM: Granite 4.1 8B IBM $0.050 $0.10 $0.062 100 0.3s 131k
IBM: Granite 4.2 8B IBM $0.10 $0.15 $0.11 92 0.4s 131k
Meta: Llama 3.1 70B Instruct Meta $0.80 $0.80 $0.80 41 0.2s 131k
Meta: Llama 3.1 8B Instruct Meta $0.22 $0.22 $0.22 102 0.2s 131k
Meta: Llama 3.3 70B Instruct Meta $0.71 $0.71 $0.71 76 0.8s 128k
MiniMax: MiniMax M2.5 MiniMax $0.30 $1.20 $0.53 78 0.5s 197k
MiniMax: MiniMax M3 MiniMax $0.23 $0.96 $0.41 112 0.4s 262k
MoonshotAI: Kimi K2.6 Moonshot AI $0.65 $3.41 $1.34 245 0.3s 262k
MoonshotAI: Kimi K2.7 Code Moonshot AI $0.71 $3.50 $1.41 94 1.2s 262k
NVIDIA: Nemotron 3.5 Lightning NVIDIA $0.10 $0.25 $0.14 140 0.1s 262k
OpenAI: gpt-oss-120b OpenAI $0.030 $0.17 $0.065 45 0.4s 131k
OpenAI: gpt-oss-20b OpenAI $0.030 $0.13 $0.055 154 0.1s 131k
Qwen: Qwen3 30B A3B Instruct 2507 Alibaba $0.10 $0.30 $0.15 45 0.3s 262k
Qwen: Qwen3 Coder 480B A35B Alibaba $1.00 $1.50 $1.12 54 0.4s 262k
Qwen: Qwen3.5-35B-A3B Alibaba $0.25 $1.25 $0.50 212 0.3s 262k
Qwen: Qwen3.6 27B Alibaba $0.60 $3.60 $1.35 60 0.4s 262k
Qwen: Qwen3.6 35B A3B Alibaba $0.25 $1.25 $0.50 80 0.6s 262k
Qwen: Qwen3.8 27B Alibaba $0.40 $3.00 $1.05 55 0.4s 262k
Z.ai: GLM 5.1 Z.ai $1.40 $4.40 $2.15 79 0.6s 203k
Z.ai: GLM 5.2 Z.ai $0.76 $2.42 $1.18 104 0.6s 1.05M
Z.ai: GLM 5.3 Flash Z.ai $0.15 $0.50 $0.24 54 1.1s 1.05M
The Known Good
Source. Prices, endpoint availability and throughput for CoreWeave are ingested from OpenRouter; model metadata is enriched from models.dev. A blank cell means the figure was not published for that endpoint — it never means zero. Blended price is (3 × input + output) ÷ 4, the same 3:1 blend used everywhere on this site; see methodology. Quality scores are not provider-specific: a model scores the same wherever it is hosted, so quality lives on the model pages.