Akash

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 Akash

Model name

12 models

Model	Context	Input /M	Output /M	Cache read /M	Cache hit	TTFT	Throughput
Gemma 4 26B A4B FP4 google/gemma-4-26b-a4b-it:akash	131k	$0.0735/M	$0.252/M	$0.0483/M	0%	1.3s	37 tps
Gemma 4 26B A4B (Thinking) FP4 google/gemma-4-26b-a4b-it:thinking:akash	131k	$0.0735/M	$0.252/M	$0.0483/M	18.8%	1.3s	37 tps
GLM 5.3 FP8 z-ai/glm-5.3:akash	1.0M	$0.20/M	$4.62/M	$0.20/M	60.3%	1.7s	82 tps
GLM 5.3 (Thinking) FP8 z-ai/glm-5.3:thinking:akash	1.0M	$0.20/M	$4.62/M	$0.20/M	16%	1.7s	82 tps
GPT OSS 120B BF16 openai/gpt-oss-120b:akash	131k	$0.0315/M	$0.196/M	$0.0315/M	N/A	1.2s	43 tps
GPT OSS 20B FP4 openai/gpt-oss-20b:akash	131k	$0.021/M	$0.105/M	N/A	N/A	0.7s	110 tps
Kimi K3 FP4 moonshotai/kimi-k3:akash	1.0M	$1.26/M	$14.70/M	$1.26/M	22.1%	2s	38.5 tps
Llama 3.3 70B Instruct FP8 meta-llama/llama-3.3-70b-instruct:akash	131k	$0.21/M	$0.546/M	$0.105/M	N/A	0.7s	36 tps
Qwen 3.6 35B A3B FP8 Qwen/Qwen3.6-35B-A3B:akash	262k	$0.105/M	$0.945/M	$0.0525/M	N/A	0.8s	82 tps
Qwen 3.6 35B A3B Thinking FP8 Qwen/Qwen3.6-35B-A3B:thinking:akash	262k	$0.105/M	$0.945/M	$0.0525/M	N/A	0.8s	82 tps
Qwen3.8 27B FP8 qwen3.8-27b:akash	262k	$0.236/M	$2.08/M	$0.0525/M	N/A	0.6s	43 tps
Qwen3.8 27B Thinking FP8 qwen3.8-27b:thinking:akash	262k	$0.236/M	$2.08/M	$0.0525/M	N/A	0.6s	43 tps

Gemma 4 26B A4B

FP4

google/gemma-4-26b-a4b-it:akash

Context

131k

Input /M

$0.0735/M

Output /M

$0.252/M

Cache read /M

$0.0483/M

Cache hit

0%

TTFT

1.3s

Throughput

37 tps

Gemma 4 26B A4B (Thinking)

FP4

google/gemma-4-26b-a4b-it:thinking:akash

Context

131k

Input /M

$0.0735/M

Output /M

$0.252/M

Cache read /M

$0.0483/M

Cache hit

18.8%

TTFT

1.3s

Throughput

37 tps

GLM 5.3

FP8

z-ai/glm-5.3:akash

Context

1.0M

Input /M

$0.20/M

Output /M

$4.62/M

Cache read /M

$0.20/M

Cache hit

60.3%

TTFT

1.7s

Throughput

82 tps

GLM 5.3 (Thinking)

FP8

z-ai/glm-5.3:thinking:akash

Context

1.0M

Input /M

$0.20/M

Output /M

$4.62/M

Cache read /M

$0.20/M

Cache hit

16%

TTFT

1.7s

Throughput

82 tps

GPT OSS 120B

BF16

openai/gpt-oss-120b:akash

Context

131k

Input /M

$0.0315/M

Output /M

$0.196/M

Cache read /M

$0.0315/M

Cache hit

N/A

TTFT

1.2s

Throughput

43 tps

GPT OSS 20B

FP4

openai/gpt-oss-20b:akash

Context

131k

Input /M

$0.021/M

Output /M

$0.105/M

Cache read /M

N/A

Cache hit

N/A

TTFT

0.7s

Throughput

110 tps

Kimi K3

FP4

moonshotai/kimi-k3:akash

Context

1.0M

Input /M

$1.26/M

Output /M

$14.70/M

Cache read /M

$1.26/M

Cache hit

22.1%

TTFT

2s

Throughput

38.5 tps

Llama 3.3 70B Instruct

FP8

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

Context

131k

Input /M

$0.21/M

Output /M

$0.546/M

Cache read /M

$0.105/M

Cache hit

N/A

TTFT

0.7s

Throughput

36 tps

Qwen 3.6 35B A3B

FP8

Qwen/Qwen3.6-35B-A3B:akash

Context

262k

Input /M

$0.105/M

Output /M

$0.945/M

Cache read /M

$0.0525/M

Cache hit

N/A

TTFT

0.8s

Throughput

82 tps

Qwen 3.6 35B A3B Thinking

FP8

Qwen/Qwen3.6-35B-A3B:thinking:akash

Context

262k

Input /M

$0.105/M

Output /M

$0.945/M

Cache read /M

$0.0525/M

Cache hit

N/A

TTFT

0.8s

Throughput

82 tps

Qwen3.8 27B

FP8

qwen3.8-27b:akash

Context

262k

Input /M

$0.236/M

Output /M

$2.08/M

Cache read /M

$0.0525/M

Cache hit

N/A

TTFT

0.6s

Throughput

43 tps

Qwen3.8 27B Thinking

FP8

qwen3.8-27b:thinking:akash

Context

262k

Input /M

$0.236/M

Output /M

$2.08/M

Cache read /M

$0.0525/M

Cache hit

N/A

TTFT

0.6s

Throughput

43 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 Akash via the API

Call a model served by Akash 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": "google/gemma-4-26b-a4b-it:akash",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

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