none reviewed yet · 5 benchmark observations · capture dates not recorded
Inspect the evidenceQwen · MODEL PROFILE
Qwen3.8-Flash-Next
Budget-conscious frontend teams and self-hosters who want frontier-adjacent UI quality
Saved editorial fit for your current task lens; not a benchmark result
How scoring worksNo verified rate for your workload
Adjust workloadAt a glance
9 facts- Provider
- Qwen
- Type
- API model
- Access
- Hugging Face weights for self-hosting; hosted `qwen3.8-flash` via Alibaba Cloud Model Studio / QwenCloud
- Context
- 262,144 tokens native, extensible to 1,000,000 via YaRN/rope edit (hosted qwen3.8-flash defaults to 1M); AA records 256k
- Maximum output
- Recommended 262,144 reasoning + 131,072 final response at 1M configuration
- License / weights
- Open weights
- Input modalities
- Causal LM with vision encoder: text, image and video in; text out
- Output modalities
- Not recorded
- Pricing notes
- Hosted qwen3.8-flash $0.15 in / $0.47 out per 1M (Singapore/international; $0.113/$0.382 other regions); cache read $0.016, cache write $0.20; self-hosting has no token price
The practical trade-offs
Documented strengths
- Best open-weight frontend result on Arena WebDev Frontend (1643, rank 8)
- Vision2Web 64.0 - a directly frontend-relevant visual web-dev score
- LiveCodeBench v6 91.9 and SWE-bench Pro 62.5% at $0.15/$0.47
- 125B total / 6B active parameters makes self-hosting comparatively cheap
Limitations to check
- Qwen Community License 1.0 restricts commercial MaaS use without a separate agreement
- Native context is 262K, not 1M - reaching 1M needs config changes when self-hosted
- The weights and the hosted API are documented as different artifacts, so benchmarks may not transfer
- Only 1,720 Arena votes
Follow the source
6 source linksEvidence & conditions
6Source links
NoneReviewed records
5Benchmark runs
Not recordedLatest capture
- Evidence source 1
huggingface.coDate not recorded - Evidence source 2
digitalapplied.comDate not recorded - Evidence source 3
alibabacloud.comDate not recorded - Evidence source 4
arena.aiDate not recorded - Evidence source 5
benchlm.aiDate not recorded - Evidence source 6
artificialanalysis.aiDate not recorded
Benchmark observations
SWE-bench Pro (provider-corrected tasks) Inspect conditions
- name
- SWE-bench Pro (provider-corrected tasks)
- value
- 62.5
- unit
- %
- source Url
- https://huggingface.co/Qwen/Qwen3.8-Flash-Next
- checked At
- 2026-09-08T19:28:18Z
- attribution
- provider-reported
- configuration
- Qwen3.8-Flash-Next; Claude Code; temperature=1; top_p=0.95; 256K context; provider corrected problematic tasks and reran baselines.
DeepSWE v1.1 Inspect conditions
- name
- DeepSWE v1.1
- value
- 58.7
- unit
- %
- source Url
- https://huggingface.co/Qwen/Qwen3.8-Flash-Next
- checked At
- 2026-09-08T19:28:18Z
- attribution
- provider-reported
- configuration
- Qwen3.8-Flash-Next; mini-SWE-agent gave the best result across evaluated harnesses; temperature=1; top_p=0.95; 256K context.
SWE-bench Multilingual Inspect conditions
- name
- SWE-bench Multilingual
- value
- 81
- unit
- %
- source Url
- https://huggingface.co/Qwen/Qwen3.8-Flash-Next
- checked At
- 2026-09-08T19:28:18Z
- attribution
- provider-reported
- configuration
- Qwen3.8-Flash-Next; mini-SWE-agent; temperature=1; top_p=0.95; 256K context.
NL2Repo Inspect conditions
- name
- NL2Repo
- value
- 48.1
- unit
- %
- source Url
- https://huggingface.co/Qwen/Qwen3.8-Flash-Next
- checked At
- 2026-09-08T19:28:18Z
- attribution
- provider-reported
- configuration
- Qwen3.8-Flash-Next; Claude Code; repository downloading via Bash disabled.
Arena WebDev Frontend Inspect conditions
- name
- Arena WebDev Frontend
- value
- 1643
- unit
- arena_score
- source Url
- https://arena.ai/leaderboard/code/webdev/frontend
- checked At
- 2026-09-08T19:28:18Z
- attribution
- benchmark-publisher
- configuration
- Exact row qwen3.8-flash-next; source results dated 2026-09-05; human preference; confidence interval +/-16; 1720 votes; preliminary. This is a checked source snapshot, not a newly run September 8 evaluation.
Source reports, measured results, and editorial assessments are separate evidence types. Source dates do not imply a new benchmark run. Capture dates are shown when recorded; otherwise we say the capture date was not recorded. Recorded dates are compared with this catalogue snapshot (Sep 11, 2026), not today’s date.
Method AIS-V2
How this page is assessed
Quality claims follow Method AIS-V2.
Money never buys rank — commercial relationships cannot change rank, eligibility, evidence quality, or publication timing.
Submit or claim a listing is free and does not skip the editorial queue.
Private · this device only