Where is the Qwen3.8 27B documentation?

The Qwen3.8 27B documentation on OpenRouter includes the Self-Hosted LLM vs API: When Does Self-Hosting Pay Off blog post and the LLM Quantization: What fp8 and fp4 Do to Quality and Price blog post. The quickstart covers API keys and your first request, and the API reference lists every parameter.

How much does Qwen3.8 27B cost?

Qwen3.8 27B costs $0.035/M input tokens and $1.50/M output tokens, with separate rates for Cache Read at $0.0175/M tokens.

What is the context length of Qwen3.8 27B?

Qwen3.8 27B has a 1,024,000 token context window. It supports up to 131,072 completion tokens.

Does Qwen3.8 27B support tool calling and structured outputs?

Yes. Qwen3.8 27B accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.

What inputs and outputs does Qwen3.8 27B support?

Qwen3.8 27B accepts text, images, and video as input and returns text.

Which providers serve Qwen3.8 27B?

Qwen3.8 27B is served by 23 providers on OpenRouter: Darkbloom, Chutes, NEAR AI, Ionstream, Modal, DekaLLM, Prime Intellect, Wafer and 15 more. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.

When was Qwen3.8 27B released?

Qwen3.8 27B was released on August 14, 2026.

Qwen: Qwen3.8 27B (nitro)

qwen/qwen3.8-27b:nitro

Model weights

Compare

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 enabled or disabled.

Modalities

In / out price

$0.035 / $1.50per 1M

Context

1.0M

Released

Aug 14, 2026

Providers

Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).

Latency / throughput

Provider	Input /M	Output /M	Cache read /M	Cache create 5m /M	Cache read 5m /M	Latency	Throughput	Uptime
	$0.99	$1.49	$0.99	--	--	0.29s	945 tok/s	100.00%
	$0.45	$3.00	$0.10	--	--	0.97s	236 tok/s	99.76%
	$0.06	$2.20	$0.059	--	--	0.43s	142 tok/s	99.68%
	$0.04	$2.30	$0.02	--	--	0.86s	120 tok/s	100.00%
	$0.40	$3.00	$0.15	--	--	0.50s	118 tok/s	99.97%
	$0.15	$1.60	$0.05	--	--	0.69s	101 tok/s	99.58%
	$0.24	$2.20	$0.05	--	--	0.72s	94 tok/s	100.00%
	$0.04	$2.50	$0.02	--	--	0.99s	89 tok/s	97.24%
	$0.04	$1.35	$0.018	--	--	1.26s	78 tok/s	99.89%
	$0.125	$2.00	$0.05	--	--	1.27s	61 tok/s	99.67%
	$0.15	$2.50	$0.05	--	--	1.63s	60 tok/s	100.00%
	$0.45	$3.20	--	--	--	0.91s	58 tok/s	99.85%
	$0.072	$0.511	$0.008	--	--	0.93s	54 tok/s	99.79%
	$0.225	$1.98	$0.05	--	--	0.73s	52 tok/s	99.80%
25% off	$0.20$0.15	$2.50$1.875	$0.05$0.0375	--	--	0.87s	51 tok/s	100.00%
	$0.42	$3.00	$0.085	--	--	6.83s	44 tok/s	100.00%
	$0.425	$2.55	$0.085	$0.53125	$0.0425	0.79s	42 tok/s	99.67%
	$0.45	$3.20	$0.05	--	--	0.35s	37 tok/s	98.49%
	$0.035	$1.50	$0.0175	--	--	2.87s	35 tok/s	99.24%
	$0.039	$1.25	$0.0039	--	--	5.32s	28 tok/s	99.26%
	$0.45	$3.20	$0.25	--	--	5.01s	27 tok/s	87.45%
	$0.04	$1.95	$0.02	--	--	3.44s	16 tok/s	97.07%
65% off	$0.50$0.175	$3.00$1.05	$0.10$0.035	--	--	24.85s	8 tok/s	95.25%

Input $0.99Output $1.49
945 tok/s100.00% uptime
Input $0.45Output $3.00
236 tok/s99.76% uptime
Input $0.06Output $2.20
142 tok/s99.68% uptime
Input $0.04Output $2.30
120 tok/s100.00% uptime
Input $0.40Output $3.00
118 tok/s99.97% uptime
Input $0.15Output $1.60
101 tok/s99.58% uptime
Input $0.24Output $2.20
94 tok/s100.00% uptime
Input $0.04Output $2.50
89 tok/s97.24% uptime
Input $0.04Output $1.35
78 tok/s99.89% uptime
Input $0.125Output $2.00
61 tok/s99.67% uptime
Input $0.15Output $2.50
60 tok/s100.00% uptime
Input $0.45Output $3.20
58 tok/s99.85% uptime
Input $0.072Output $0.511
54 tok/s99.79% uptime
Input $0.225Output $1.98
52 tok/s99.80% uptime
Input $0.15Output $1.875
25% off
51 tok/s100.00% uptime
Input $0.42Output $3.00
44 tok/s100.00% uptime
Input $0.425Output $2.55
42 tok/s99.67% uptime
Input $0.45Output $3.20
37 tok/s98.49% uptime
Input $0.035Output $1.50
35 tok/s99.24% uptime
Input $0.039Output $1.25
28 tok/s99.26% uptime
Input $0.45Output $3.20
27 tok/s87.45% uptime
Input $0.04Output $1.95
16 tok/s97.07% uptime
Input $0.175Output $1.05
65% off
8 tok/s95.25% uptime

Pricing

The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.

Weighted average

Input

$0.06917/M tokens

Output

$2.084/M tokens

Qwen3.8 27B — Price history

Price history

No effective price data for this selection

$0.0130	$1.249	74.1%	4.5%
$0.0222	$2.299	89.0%	23.5%
$0.0229	$1.349	77.5%	10.1%
$0.0243	$2.500	78.3%	5.5%
$0.0273	$1.949	63.3%	0.5%
$0.0327	$1.500	12.2%	0.9%
$0.0593	$2.200	67.4%	15.2%
$0.0618	$1.874	78.4%	6.0%
$0.0755	$1.600	74.4%	2.1%
$0.0788	$0.853	24.7%	1.1%
$0.0800	$2.199	84.2%	4.1%
$0.0917	$1.999	44.3%	0.4%
$0.0924	$1.049	59.0%	2.6%
$0.1294	$3.000	91.6%	15.6%
$0.1336	$2.500	16.3%	1.5%
$0.1645	$2.550	76.7%	3.6%
$0.1695	$1.978	31.6%	0.6%
$0.1919	$3.000	68.1%	0.5%
$0.2747	$3.000	50.1%	0.9%
$0.2809	$3.200	84.5%	0.0%
$0.4498	$3.200	0.3%	0.0%
$0.4499	$3.199	0.0%	0.0%
$0.9899	$1.489	26.4%	0.6%

Effective in $0.0130Effective out $1.249
Token share
4.5%
Effective in $0.0222Effective out $2.299
Token share
23.5%
Effective in $0.0229Effective out $1.349
Token share
10.1%
Effective in $0.0243Effective out $2.500
Token share
5.5%
Effective in $0.0273Effective out $1.949
Token share
0.5%
Effective in $0.0327Effective out $1.500
Token share
0.9%
Effective in $0.0593Effective out $2.200
Token share
15.2%
Effective in $0.0618Effective out $1.874
Token share
6.0%
Effective in $0.0755Effective out $1.600
Token share
2.1%
Effective in $0.0788Effective out $0.853
Token share
1.1%
Effective in $0.0800Effective out $2.199
Token share
4.1%
Effective in $0.0917Effective out $1.999
Token share
0.4%
Effective in $0.0924Effective out $1.049
Token share
2.6%
Effective in $0.1294Effective out $3.000
Token share
15.6%
Effective in $0.1336Effective out $2.500
Token share
1.5%
Effective in $0.1645Effective out $2.550
Token share
3.6%
Effective in $0.1695Effective out $1.978
Token share
0.6%
Effective in $0.1919Effective out $3.000
Token share
0.5%
Effective in $0.2747Effective out $3.000
Token share
0.9%
Effective in $0.2809Effective out $3.200
Token share
0.0%
Effective in $0.4498Effective out $3.200
Token share
0.0%
Effective in $0.4499Effective out $3.199
Token share
0.0%
Effective in $0.9899Effective out $1.489
Token share
0.6%

Performance

Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).

Throughput

945tok/s

P50, best across providers

Latency

0.29s

P50, best provider

Reasoning effortLatency / throughput

AutoExacto benchmarks

GPQA DiamondTAU-BenchStreamLake84.9%--GMICloud82.6%--DekaLLM83.4%77.9%Parasail83.9%77.3%Prime Intellect83.2%78.0%

+21 more providers

Uptime

Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.

Uptime (3d)

100.00%

Availability (3d)

99.87%

Availability over the last 3 days

Last 72 hours

Availability 99.87%

3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability

99.91%

Without Routing

96.07%

When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.

Benchmarks

Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.

Benchmark score summary for Qwen: Qwen3.8 27B (Artificial Analysis and Vals AI)
Source	Benchmark	Score
Artificial Analysis	Qwen3.8 27B (Non-reasoning) Intelligence Index	10.0
Artificial Analysis	Qwen3.8 27B (Non-reasoning) Coding Index	44.6
Artificial Analysis	Qwen3.8 27B (Non-reasoning) Agentic Index	22.4
Artificial Analysis	Qwen3.8 27B (Non-reasoning) GPQA Diamond	81.8%
Artificial Analysis	Qwen3.8 27B (Non-reasoning) HLE	12.1%
Artificial Analysis	Qwen3.8 27B (Non-reasoning) AA-LCR	69.3%
Artificial Analysis	Qwen3.8 27B (Non-reasoning) τ-Bench Banking	20.0%
Artificial Analysis	Qwen3.8 27B (Non-reasoning) GDPval-AA	29.1%
Artificial Analysis	Qwen3.8 27B (Non-reasoning) CritPt	0.3%
Artificial Analysis	Qwen3.8 27B (Non-reasoning) SciCode	36.2%
Artificial Analysis	Qwen3.8 27B (Non-reasoning) Terminal-Bench 4.0	0.0%
Artificial Analysis	Qwen3.8 27B (Non-reasoning) Terminal-Bench 2.1	49.1%
Artificial Analysis	Qwen3.8 27B (Non-reasoning) AA-Omniscience Accuracy	9.1%
Artificial Analysis	Qwen3.8 27B (Non-reasoning) AA-Omniscience Non-Hallucination Rate	82.0%
Artificial Analysis	Qwen3.8 27B (Xhigh) Intelligence Index	20.1
Artificial Analysis	Qwen3.8 27B (Xhigh) Coding Index	68.1
Artificial Analysis	Qwen3.8 27B (Xhigh) Agentic Index	45.8
Artificial Analysis	Qwen3.8 27B (Xhigh) GPQA Diamond	90.5%
Artificial Analysis	Qwen3.8 27B (Xhigh) HLE	33.9%
Artificial Analysis	Qwen3.8 27B (Xhigh) AA-LCR	82.0%
Artificial Analysis	Qwen3.8 27B (Xhigh) τ-Bench Banking	48.0%
Artificial Analysis	Qwen3.8 27B (Xhigh) GDPval-AA	46.3%
Artificial Analysis	Qwen3.8 27B (Xhigh) CritPt	5.4%
Artificial Analysis	Qwen3.8 27B (Xhigh) SciCode	46.6%
Artificial Analysis	Qwen3.8 27B (Xhigh) Terminal-Bench 4.0	5.6%
Artificial Analysis	Qwen3.8 27B (Xhigh) Terminal-Bench 2.1	79.8%
Artificial Analysis	Qwen3.8 27B (Xhigh) AA-Omniscience Accuracy	17.3%
Artificial Analysis	Qwen3.8 27B (Xhigh) AA-Omniscience Non-Hallucination Rate	71.4%
Artificial Analysis	Qwen3.8 27B (Low) Intelligence Index	15.7
Artificial Analysis	Qwen3.8 27B (Low) Coding Index	58.2
Artificial Analysis	Qwen3.8 27B (Low) Agentic Index	38.9
Artificial Analysis	Qwen3.8 27B (Low) GPQA Diamond	84.5%
Artificial Analysis	Qwen3.8 27B (Low) HLE	14.0%
Artificial Analysis	Qwen3.8 27B (Low) AA-LCR	77.3%
Artificial Analysis	Qwen3.8 27B (Low) τ-Bench Banking	32.2%
Artificial Analysis	Qwen3.8 27B (Low) GDPval-AA	43.3%
Artificial Analysis	Qwen3.8 27B (Low) CritPt	0.0%
Artificial Analysis	Qwen3.8 27B (Low) SciCode	40.0%
Artificial Analysis	Qwen3.8 27B (Low) Terminal-Bench 4.0	2.5%
Artificial Analysis	Qwen3.8 27B (Low) Terminal-Bench 2.1	67.4%
Artificial Analysis	Qwen3.8 27B (Low) AA-Omniscience Accuracy	18.3%
Artificial Analysis	Qwen3.8 27B (Low) AA-Omniscience Non-Hallucination Rate	51.3%
Artificial Analysis	Qwen3.8 27B (Medium) Intelligence Index	17.5
Artificial Analysis	Qwen3.8 27B (Medium) Coding Index	56.1
Artificial Analysis	Qwen3.8 27B (Medium) Agentic Index	44.4
Artificial Analysis	Qwen3.8 27B (Medium) GPQA Diamond	84.5%
Artificial Analysis	Qwen3.8 27B (Medium) HLE	14.1%
Artificial Analysis	Qwen3.8 27B (Medium) AA-LCR	79.7%
Artificial Analysis	Qwen3.8 27B (Medium) τ-Bench Banking	47.4%
Artificial Analysis	Qwen3.8 27B (Medium) GDPval-AA	44.5%
Artificial Analysis	Qwen3.8 27B (Medium) CritPt	0.0%
Artificial Analysis	Qwen3.8 27B (Medium) SciCode	39.0%
Artificial Analysis	Qwen3.8 27B (Medium) Terminal-Bench 4.0	5.1%
Artificial Analysis	Qwen3.8 27B (Medium) Terminal-Bench 2.1	65.2%
Artificial Analysis	Qwen3.8 27B (Medium) AA-Omniscience Accuracy	17.9%
Artificial Analysis	Qwen3.8 27B (Medium) AA-Omniscience Non-Hallucination Rate	34.6%
Vals AI	Qwen 3.8 27B Code Migration	14.2%
Vals AI	Qwen 3.8 27B CyberBench v1.1	67.8%
Vals AI	Qwen 3.8 27B EMB	59.7%
Vals AI	Qwen 3.8 27B Finance Agent (v2)	48.6%
Vals AI	Qwen 3.8 27B GPQA Diamond	88.9%
Vals AI	Qwen 3.8 27B Harvey's Legal Agent Benchmark v1	11.3%
Vals AI	Qwen 3.8 27B IOI	39.1%
Vals AI	Qwen 3.8 27B LiveCodeBench	84.0%
Vals AI	Qwen 3.8 27B LegalBench	82.4%
Vals AI	Qwen 3.8 27B Legal Research Bench	36.1%
Vals AI	Qwen 3.8 27B MedCode	28.7%
Vals AI	Qwen 3.8 27B MedScribe	83.8%
Vals AI	Qwen 3.8 27B MMLU Pro	84.3%
Vals AI	Qwen 3.8 27B MMMU Pro	83.9%
Vals AI	Qwen 3.8 27B MortgageTax	64.9%
Vals AI	Qwen 3.8 27B ProgramBench	0.0%
Vals AI	Qwen 3.8 27B ProofBench v1.1	16.0%
Vals AI	Qwen 3.8 27B SAGE v1	52.4%
Vals AI	Qwen 3.8 27B SkillsBench	38.1%
Vals AI	Qwen 3.8 27B SWE-bench	86.0%
Vals AI	Qwen 3.8 27B Tax Agent Bench	30.9%
Vals AI	Qwen 3.8 27B TaxEval v2	70.8%
Vals AI	Qwen 3.8 27B Terminal-Bench 2.1	58.4%
Vals AI	Qwen 3.8 27B Terminal-Bench-Science	0.0%
Vals AI	Qwen 3.8 27B Vals Index	48.5%
Vals AI	Qwen 3.8 27B Vibe Code Bench v1.1	64.8%

Apps

Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.

Portkey AI

Control panel for AI apps

199Btokens

Codex

A coding agent that helps you build and ship with AI

187Btokens

Hermes Agent

Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.

106Btokens

Claude Code

Claude Code is Anthropic's agentic coding tool that reads your entire codebase, plans and executes changes across files, runs tests, and iterates on failures, all from natural language prompts.

83.4Btokens

There are many coding agents, but this one is yours.

74.4Btokens

Rest of OpenRouter

38B

Top 5 Apps

7.78B

Not currently available through EU in-region routing. In-region routing docs

More models from Qwen

Qwen Image 2.1 Turbo

Qwen Image 2.1 Turbo is the fast, lower-cost tier of the Qwen Image 2.1 series from Qwen, for image generation and editing where throughput and price matter more than peak fidelity. See Qwen Image 2.1 Pro for the higher-quality tier.

Imagefrom $0.0184/image

Qwen Image 2.1 Pro

Qwen Image 2.1 Pro is an image generation and editing model from Qwen, the higher-quality tier of the Qwen Image 2.1 series. Choose it over Qwen Image 2.1 Turbo when output fidelity matters more than cost or speed.

Imagefrom $0.046/image

Qwen3.8 Max Prime

Qwen3.8 Max Prime is a higher-throughput variant of Qwen3.8 Max from Alibaba's Qwen team, served as a separate SKU at a higher price point. It accepts text, image, and video input and returns text, with a 1M-token context window and reasoning enabled by default. Tool calling, structured outputs, and configurable reasoning effort are supported, matching Qwen3.8 Max.

Text1M context$4 / $12

Qwen3.8 Omni Flash

Qwen3.8 Omni Flash is an omni-modal reasoning model from Alibaba, the first Qwen model built around agentic capabilities with native audio-video understanding. It is suited for audio-video analysis and summarization, video editing and production workflows, audio-video dialogue, coding, knowledge work, and GUI interaction, and it is particularly strong at long-form multimedia tasks that combine speech, sound, and visual context. It also supports two-channel and four-channel spatial audio understanding.

Text1M context$0.15 / $0.47

Qwen3.8 Max

Qwen3.8 Max 0902 is an updated snapshot of Qwen3.8 Max from Alibaba's Qwen team. It is a 2.4-trillion-parameter mixture-of-experts model that accepts text, image, and video input and returns text, with a 1M-token context window and reasoning enabled by default.

This snapshot is post-trained for coding and agentic work, including multi-step software projects, multi-tool orchestration, and long-horizon task execution. It also targets chart reasoning, document parsing, and multimodal understanding over long documents and extended video. Tool calling, structured outputs, and configurable reasoning effort are supported.

Text1M context$2 / $6

Qwen3.8 Max (batch)

Text262K context$1.15 / $3.45

Qwen3.8 Flash

Qwen3.8 Flash is a multimodal reasoning model from Alibaba. It is suited for coding assistance, agentic workflows, visual understanding, document and codebase analysis, desktop interaction, chart analysis, and long-video analysis.

Text1M context$0.15 / $0.47

Qwen3 Reranker 0.6B

Qwen3 Reranker 0.6B is the smallest text reranking model from Alibaba Cloud in the Qwen3 Reranker series. Built on the Qwen3 architecture, it evaluates query-document pairs to produce relevance scores for retrieval and RAG pipelines. Supports 100+ languages with instruction-aware reranking. Designed for low-latency, resource-constrained deployments.

Rerank$0.20/M tokens

Qwen3 Reranker 4B

Qwen3 Reranker 4B is a text reranking model from Alibaba Cloud built on the Qwen3 architecture. It evaluates query-document pairs to produce relevance scores for use in retrieval and RAG pipelines. Supports 100+ languages and programming languages, with instruction-aware reranking that allows customizing scoring criteria per task. A compact 4B alternative balancing performance and inference cost.

Rerank

Qwen3.8 27B

Text1.0M context$0.035 / $1.50

Qwen3.8 27B (batch)

Text262K context$0.16 / $2

Qwen3 Reranker 8B

Qwen3 Reranker 8B is a text reranking model from Alibaba Cloud built on the Qwen3 architecture. It evaluates query-document pairs to produce relevance scores for use in retrieval and RAG pipelines. Supports 100+ languages and programming languages, with instruction-aware reranking that allows customizing scoring criteria per task. Offers strong performance on multilingual benchmarks including MTEB, CMTEB, and MMTEB.

Rerank$0.20/M tokens

Qwen3 ASR 1.7B

Qwen3 ASR 1.7B is an automatic speech recognition model from Qwen. It supports multilingual language identification and transcription across 30 languages and 22 Chinese dialects, with streaming and offline inference plus segment-level and word-level timestamps.

Transcription$0.000008/second

Qwen3 ASR 0.6B

Qwen3 ASR 0.6B is a compact automatic speech recognition model from Qwen. It supports multilingual language identification and transcription across 30 languages and 22 Chinese dialects, with streaming and offline inference plus segment-level and word-level timestamps.

Transcription$0.000003/second

Qwen3.8 2.4T A95B

Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of Qwen3.8 Max, with 95 billion active parameters out of 2.4 trillion total. It is suited for coding, research, complex reasoning, and agentic workflows.

Text1.0M context$2 / $6

Qwen Image 3 Pro

Qwen Image 3 Pro is an image generation and editing model from Qwen. It supports precise rendering of text and details as small as 10px, along with richer world knowledge compared to previous generations.

Imagefrom $0.04/image

Qwen Image 3

Qwen Image 3 is a unified image generation and editing model from Qwen. It supports precise rendering of text and details as small as 10px, along with a richer world knowledge base than previous generations.

Imagefrom $0.03/image

Qwen3.8 Max

Qwen3.8 Max (0803) is the August 3, 2026 checkpoint of Qwen3.8 Max, the flagship model in Alibaba's Qwen3.8 series and the general-availability successor to the Qwen3.8 Max Preview. It is a multimodal reasoning model intended for complex reasoning, visual understanding, coding, and agentic workflows. This checkpoint was superseded by Qwen3.8 Max (0902) on September 5, 2026.

Text1M context

Qwen3.7 Flash

Qwen3.7 Flash is a vision-language reasoning model from Alibaba. It is suited for multimodal agents, visual coding, search, and computer interaction, with strengths in object recognition, spatial understanding, and real-world visual perception.

Text1M context$0.03 / $0.13

Qwen-Audio-3.0-TTS Flash

Qwen-Audio-3.0-TTS Flash is Alibaba's fast, cost-efficient text-to-speech model, generating spoken audio from text via the DashScope Speech Synthesizer API.

Speech$15/M characters

Qwen: Qwen3.8 27B (nitro)

Providers