The Known Good Updated 8 Oct 2026 Subscribe
Providers· Inference host· 14 tracked models

PrimeIntellect — pricing & performance

Compare models All providers Data sources
Models served
14
of 1,191 tracked models
Mean output speed
74
tokens per second, averaged over
14 measured endpoints
Mean latency
1.1s
time to first token, averaged over
14 measured endpoints
Cheapest blended
$0.13
per 1M tokens at a 3:1
input:output blend
Output speed on PrimeIntellect

Output tokens per second for all 14 models on PrimeIntellect with published throughput telemetry, the most recent figure OpenRouter publishes for each host. Higher is better. We do not run these measurements ourselves.

⇩
170
130
124
84
80
66
56
55
54
53
50
39
The Known Good
14 models on PrimeIntellect
Full leaderboard →
Model Creator Input $/M Output $/M Blended Speed TTFT Context
Meta: Llama 3.3 70B Instruct Meta $1.04 $2.25 $1.34 38 0.7s 131k
MiniMax: MiniMax M2.5 MiniMax $0.60 $2.40 $1.05 56 1.4s 205k
MiniMax: MiniMax M2.7 MiniMax $0.60 $2.40 $1.05 50 1.4s 205k
Mistral: Mistral Large 3 2512 Mistral $0.50 $1.50 $0.75 55 1.1s 262k
Mistral: Mistral Large 4 Mistral $0.68 $2.09 $1.03 35 2.6s 524k
OpenAI: gpt-oss-20b OpenAI $0.070 $0.30 $0.13 170 0.6s 131k
Qwen: Qwen3 235B A22B Instruct 2507 Alibaba $0.25 $1.00 $0.44 53 0.7s 262k
Qwen: Qwen3 30B A3B Instruct 2507 Alibaba $0.13 $0.52 $0.23 54 0.5s 262k
Qwen: Qwen3 Coder Next Alibaba $0.30 $1.50 $0.60 124 1.5s 262k
Qwen: Qwen3 VL 235B A22B Instruct Alibaba $0.30 $1.50 $0.60 39 1.3s 262k
Qwen: Qwen3 VL 30B A3B Instruct Alibaba $0.20 $0.70 $0.33 84 0.9s 262k
Qwen: Qwen3 VL 8B Instruct Alibaba $0.12 $0.46 $0.20 66 0.9s 262k
Z.ai: GLM 5.3 Z.ai $1.40 $4.40 $2.15 130 1.1s 1.05M
Z.ai: GLM 5.3 Flash Z.ai $0.15 $0.50 $0.24 80 1.2s 1.05M
The Known Good
Source. Prices, endpoint availability and throughput for PrimeIntellect 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.