PrimeIntellect

Prime Intellect offers both hosted and externally routed inference. Its public policies do not establish prompt retention, training guarantees, or processing locations for externally routed inference; do not assume zero retention or exclusion from training.

Data handling can vary by model and deployment. The privacy badge above shows the more conservative of the provider-wide fallback and the policies for routes on this page; each model page shows the policy for that specific route.

Models served by PrimeIntellect

Model name

Model	Context	Input /M	Output /M	TTFT	Throughput
GLM 5.3 Flash z-ai/glm-5.3-flash:primeintellect	1.0M	$0.16/M	$0.53/M	1.7s	70 tps
GPT OSS 20B openai/gpt-oss-20b:primeintellect	131k	$0.07/M	$0.31/M	0.5s	139 tps
Llama 3.3 70B Instruct meta-llama/llama-3.3-70b-instruct:primeintellect	131k	$1.09/M	$2.37/M	0.4s	44.5 tps
MiniMax M2.5 minimax/minimax-m2.5:primeintellect	205k	$0.63/M	$2.52/M	1.3s	39 tps
MiniMax M2.7 minimax/minimax-m2.7:primeintellect	205k	$0.63/M	$2.52/M	1.6s	37.5 tps
Qwen3 235B A22B qwen/qwen3-235b-a22b:primeintellect	262k	$0.26/M	$1.05/M	0.7s	56 tps
Qwen3 Coder Next qwen/qwen3-coder-next:primeintellect	262k	$0.31/M	$1.57/M	2.1s	86 tps

GLM 5.3 Flash

z-ai/glm-5.3-flash:primeintellect

Context

1.0M

Input /M

$0.16/M

Output /M

$0.53/M

TTFT

1.7s

Throughput

70 tps

GPT OSS 20B

openai/gpt-oss-20b:primeintellect

Context

131k

Input /M

$0.07/M

Output /M

$0.31/M

TTFT

0.5s

Throughput

139 tps

Llama 3.3 70B Instruct

meta-llama/llama-3.3-70b-instruct:primeintellect

Context

131k

Input /M

$1.09/M

Output /M

$2.37/M

TTFT

0.4s

Throughput

44.5 tps

MiniMax M2.5

minimax/minimax-m2.5:primeintellect

Context

205k

Input /M

$0.63/M

Output /M

$2.52/M

TTFT

1.3s

Throughput

39 tps

MiniMax M2.7

minimax/minimax-m2.7:primeintellect

Context

205k

Input /M

$0.63/M

Output /M

$2.52/M

TTFT

1.6s

Throughput

37.5 tps

Qwen3 235B A22B

qwen/qwen3-235b-a22b:primeintellect

Context

262k

Input /M

$0.26/M

Output /M

$1.05/M

TTFT

0.7s

Throughput

56 tps

Qwen3 Coder Next

qwen/qwen3-coder-next:primeintellect

Context

262k

Input /M

$0.31/M

Output /M

$1.57/M

TTFT

2.1s

Throughput

86 tps

Prices are per million tokens. TTFT is the observed time to first token. Selectable routes include the applicable provider-selection markup; fixed routes show the standard model price. Availability and pricing refresh continuously; each model page shows the live provider comparison.

Using PrimeIntellect via the API

Call a model served by PrimeIntellect using its displayed model ID, including the provider suffix when selection is supported.

curl https://nano-gpt.com/api/v1/chat/completions \
  -H "Authorization: Bearer $NANOGPT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "z-ai/glm-5.3-flash:primeintellect",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

For models with provider selection, append :primeintellect to the model ID, set "provider": "primeintellect" in the request body, or send an X-Provider: primeintellect header. Explicit provider selection adds a route-specific markup over that provider's base price; the prices in the table above already include it. Models without provider selection use the displayed model ID without a provider suffix or selection surcharge.

See the API documentation for provider preferences, price-aware routing, and error behavior.