Greenference

10 models available

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 Greenference

10 models

Model	Context	Input /M	Output /M	Cache read /M	Cache write /M	Cache hit	TTFT	Throughput
Gemma 4 31B FP8	262k	$0.16/M	$0.44/M	$0.06/M	Free	56.2%	5.6s	15.9 tps
GLM 4.7 Flash FP8	131k	$0.07/M	$0.43/M	$0.02/M	Free	7.2%	0.3s	136.3 tps
GLM 4.7 Flash (Thinking) FP8	131k	$0.07/M	$0.43/M	$0.02/M	Free	N/A	N/A	N/A
GLM 5.3 Flash FP8	1.0M	$0.15/M	$0.51/M	$0.03/M	Free	96.5%	3.8s	183.3 tps
GPT OSS 20B	131k	$0.0095/M	$0.05/M	$0.0047/M	Free	N/A	N/A	N/A
Qwen 3.6 27B INT4	262k	$0.16/M	$1.42/M	$0.08/M	Free	N/A	N/A	N/A
Qwen3 14B INT4	41k	$0.10/M	$0.30/M	$0.05/M	N/A	N/A	3.6s	N/A
Qwen3 32B INT4	41k	$0.10/M	$0.30/M	$0.05/M	Free	N/A	N/A	N/A
Qwen3 8B FP8	41k	$0.47/M	$0.47/M	$0.23/M	Free	N/A	N/A	N/A
Qwen3.5 9B NVFP4	262k	$0.04/M	$0.07/M	$0.02/M	Free	N/A	0.4s	107.2 tps

Gemma 4 31B

FP8

Context: 262k
Input /M: $0.16/M
Output /M: $0.44/M
Cache read /M: $0.06/M
Cache write /M: Free
Cache hit: 56.2%
TTFT: 5.6s
Throughput: 15.9 tps

GLM 4.7 Flash

FP8

Context: 131k
Input /M: $0.07/M
Output /M: $0.43/M
Cache read /M: $0.02/M
Cache write /M: Free
Cache hit: 7.2%
TTFT: 0.3s
Throughput: 136.3 tps

GLM 4.7 Flash (Thinking)

FP8

Context: 131k
Input /M: $0.07/M
Output /M: $0.43/M
Cache read /M: $0.02/M
Cache write /M: Free
Cache hit: N/A
TTFT: N/A
Throughput: N/A

GLM 5.3 Flash

FP8

Context: 1.0M
Input /M: $0.15/M
Output /M: $0.51/M
Cache read /M: $0.03/M
Cache write /M: Free
Cache hit: 96.5%
TTFT: 3.8s
Throughput: 183.3 tps

GPT OSS 20B

Context: 131k
Input /M: $0.0095/M
Output /M: $0.05/M
Cache read /M: $0.0047/M
Cache write /M: Free
Cache hit: N/A
TTFT: N/A
Throughput: N/A

Qwen 3.6 27B

INT4

Context: 262k
Input /M: $0.16/M
Output /M: $1.42/M
Cache read /M: $0.08/M
Cache write /M: Free
Cache hit: N/A
TTFT: N/A
Throughput: N/A

Qwen3 14B

INT4

Context: 41k
Input /M: $0.10/M
Output /M: $0.30/M
Cache read /M: $0.05/M
Cache write /M: N/A
Cache hit: N/A
TTFT: 3.6s
Throughput: N/A

Qwen3 32B

INT4

Context: 41k
Input /M: $0.10/M
Output /M: $0.30/M
Cache read /M: $0.05/M
Cache write /M: Free
Cache hit: N/A
TTFT: N/A
Throughput: N/A

Qwen3 8B

FP8

Context: 41k
Input /M: $0.47/M
Output /M: $0.47/M
Cache read /M: $0.23/M
Cache write /M: Free
Cache hit: N/A
TTFT: N/A
Throughput: N/A

Qwen3.5 9B

NVFP4

Context: 262k
Input /M: $0.04/M
Output /M: $0.07/M
Cache read /M: $0.02/M
Cache write /M: Free
Cache hit: N/A
TTFT: 0.4s
Throughput: 107.2 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 Greenference via the API

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

cURL

curl https://nano-gpt.com/api/v1/chat/completions \
  -H "Authorization: Bearer $NANOGPT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemma-4-31b-it:greenference",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

For models with provider selection, append :greenference to the model ID, set "provider": "greenference" in the request body, or send an X-Provider: greenference 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.

Greenference terms of use Greenference privacy policy