Skip to content
LLMs
Qwen logo

Qwen: Qwen3.8 27B

by Qwen

Qwen: Qwen3.8 27B is a large language model from Qwen. It costs $0.240 per million input tokens and $2.20 per million output tokens. Its context window is 1M tokens.

ReasoningTool callingStructured outputPrompt cachingOpen weightsimage inputvideo input
Input / 1M tokens
$0.240
Output / 1M tokens
$2.20
Cached input / 1M
$0.050

On repeated prefixes

Context window
1M

tokens

Who serves it cheapest

20 hosts serve Qwen3.8 27B. Same weights, same API - the price difference is pure margin and routing.

Providers serving Qwen: Qwen3.8 27B, cheapest first
ProviderInput / 1MOutput / 1MContextThroughputUptime 24h
TenstorrentCheapestfp4$0.020$0.500262K-99.5%
Chutesfp8$0.039$1.25262K-99.9%
Near AIfp8$0.040$1.35262K-100.0%
Darkbloomfp4$0.035$2.20262K-97.1%
Rekafp8$0.040$2.20262K-100.0%
DeepInfrabf16$0.150$1.88262K-76.5%
Wafer$0.040$2.30262K-100.0%
Ionstreamfp8$0.089$2.35262K-99.9%
DekaLLM$0.040$2.50262K-100.0%
AkashMLfp8$0.225$1.98262K-100.0%
Parasailfp8$0.240$2.20262K-100.0%
ModelRunfp4$0.200$2.50262K-100.0%
Mancer 2fp8$0.200$2.50262K-99.8%
Phala$0.200$2.501M-97.8%
Alibaba$0.425$2.551M-99.9%
CoreWeavefp8$0.400$3.00262K-99.8%
Novita$0.420$3.001M-99.8%
Cerebrasfp16$0.990$1.4966K-100.0%
Cloudflare$0.450$3.20262K-98.4%
Venicefp8$0.450$3.20262K-100.0%

The spread between Tenstorrent and Venice is 8.1× for identical weights. Quantization and context limits differ, so check both columns before switching.

Benchmarks

Independent scores published alongside the catalogue.

Intelligence index
33.7
Coding index
68.1
Agentic index
45.8

About Qwen3.8 27B

Qwen3.8 27B is an open-weight dense vision-language model from Qwen. It is suited for coding, professional workflows, research, multimodal interaction, and long-running agent tasks, with flexible thinking that can be...

Specifications

Qwen: Qwen3.8 27B specifications
Model IDqwen/qwen3.8-27b
ProviderQwen
Context window1M tokens
Max output236K tokens
Input modalitiestext, image, video
Output modalitiestext
Knowledge cutoff-
Open weightsYes - Qwen/Qwen3.8-27B
ReleasedAugust 14, 2026

Cheaper alternatives

Models that cost less than Qwen3.8 27B while keeping at least half its context window and every input modality it supports.

Qwen3.5 Plus 2026-04-20

$0.300 in · $1.80 out

1.1× cheaper

GLM 5.3 FlashX

$0.370 in · $1.25 out

1.2× cheaper

Qwen3.5 Plus 2026-02-15

$0.260 in · $1.56 out

1.2× cheaper

Gemini 3.1 Flash Lite

$0.250 in · $1.50 out

1.3× cheaper

Gemini 3.1 Flash Lite Preview

$0.250 in · $1.50 out

1.3× cheaper

Frequently asked

How much does Qwen: Qwen3.8 27B cost?

$0.240 per million input tokens and $2.20 per million output tokens. Cached input reads cost $0.050 per million tokens.

What is the context window of Qwen: Qwen3.8 27B?

1M tokens, with up to 236K tokens of output per request.

Which provider serves Qwen: Qwen3.8 27B cheapest?

Tenstorrent at $0.020 per million input tokens - 8.1× cheaper than Venice, the most expensive of the 20 hosts serving it.

Confirm against the source: Qwen official pricing.