Data as of October 29, 2026 · How the score is built
Released Sep 18, 20261M contextQwen3.8-Omni-Flash launch post
Qwen3.8-Omni-Flash
Decision readingQwen3.8-Omni-Flash is tracked, but not publicly ranked yet. The profile exposes 19 sourced benchmark rows and leaves unsupported fields blank until a published record exists.
Released Sep 18, 2026 — see all recent releases
Voice benchmark evidence
Qwen3.8-Omni-Flash omni evaluation is an owner-defined voice evaluation. Its result stays separate from BenchLM's weighted text-model ranking.
Decision snapshot
Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.
Capability
Unranked
field median 50.5Not eligible for a public rank
Price
API rate not published
input median $0.75No comparable first-party hosted token rate
Speed
Not measured
field median 92 tok/sTime to first token not measured
Context
1Mtokens
field median 256,000Maximum output length is tracked separately
Strongest published evidence
Instruction Following ranks #23. A well-rounded choice across a range of tasks.
Validate before choosing
19 published rows leave some tracked benchmark slots empty. No comparable first-party API token rate is published.
Source-linked · 19 displayable benchmark rows
Category score record
Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticWeight 22%2 benchmarksVerified | Score pending | 2 benchmarks | Verified | |||
| CodingRank Not rankedWeight 20%5 benchmarksVerified | 58.0 | 5 benchmarks | Verified | |||
| ReasoningWeight 17%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| MultimodalRank Not rankedWeight 12%8 benchmarksVerified | 86.2 | 8 benchmarks | Verified | |||
| KnowledgeRank Not rankedWeight 12%3 benchmarksVerified | 59.3 | 3 benchmarks | Verified | |||
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| Inst. FollowingRank #23 of 157Percentile 86thWeight 5%1 benchmarkVerified | 91.6 | 1 benchmark | Verified | |||
| MathWeight 5%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured |
19 of 782 tracked benchmark slots have displayable evidence · bars run 0–100
Coverage detailsHow much of this is verified
Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.
- Agentic2/2 verified
- Coding5/5 verified
- ReasoningNot measured
- Multimodal8/8 verified
- Knowledge3/3 verified
- MultilingualNot measured
- Inst. Following1/1 verified
- MathNot measured
Benchmark ledger
Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.
Coding5 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| SWE-bench Pro | Score63.3% | Versus best verified row Best verified: Claude Opus 5.5 · 89.9% | Gap26.6 behind | Weight26% ref. weight | Provider exact |
| DeepSWE | Score57.8% | Versus best verified row Best verified: Muse Wonder Full · 79.0% | Gap21.2 behind | Weight15% ref. weight | Provider exact |
| SWE Multilingual | Score80.5% | Versus best verified row Best verified: Claude Opus 5.5 · 93.9% | Gap13.4 behind | Weight5% ref. weight | Provider exact |
| NL2Repo | Score48.9% | Versus best verified row Best verified: DeepSeek V4.1 Flash 1017 · 68.2% | Gap19.3 behind | WeightDisplay only | Provider exact |
| LiveCodeBench v6 | Score92.6% | Versus best verified row Best verified: Sakana Fugu-Ultra · 93.2% | Gap0.6 behind | WeightDisplay only | Provider exact |
Agentic2 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| CoWorkBench | Score75.3% | Versus best verified row Best verified: Qwen3.8-Omni-Flash · 75.3% | GapBest verified | WeightDisplay only | Provider exact |
| AndroidWorld | Score87.1% | Versus best verified row Best verified: Qwen3.8-Omni-Flash · 87.1% | GapBest verified | WeightDisplay only | Provider exact |
Multimodal8 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| CharXivCharXiv Reasoning | Score91.4% | Versus best verified row Best verified: Qwen3.8 Max · 93.5% | Gap2.1 behind | WeightWeighted 20% | Provider exact |
| Vision2Web | Score62.9% | Versus best verified row Best verified: Qwen3.8 Max · 69.0% | Gap6.1 behind | WeightDisplay only | Provider exact |
| ERQA | Score71.0% | Versus best verified row Best verified: Qwen3.8 Max · 77.8% | Gap6.8 behind | WeightDisplay only | Provider exact |
| LVBench | Score76.9% | Versus best verified row Best verified: Gemini 4 Argon · 91.7% | Gap14.8 behind | WeightDisplay only | Provider exact |
| RealWorldQA | Score87.7% | Versus best verified row Best verified: Qwen3.8-Flash-Next · 88.5% | Gap0.8 behind | WeightDisplay only | Provider exact |
| MathVision | Score91.8% | Versus best verified row Best verified: Qwen3.8 Max · 95.2% | Gap3.4 behind | WeightDisplay only | Provider exact |
| MathVision w/ PythonMathVision with Python | Score96.2% | Versus best verified row Best verified: Kimi K3 · 97.8% | Gap1.6 behind | WeightDisplay only | Provider exact |
| CharXiv w/o toolsCharXiv Reasoning without tools | Score83.5% | Versus best verified row Best verified: Kimi K3.1 Pro · 89% | Gap5.5 behind | WeightDisplay only | Provider exact |
Knowledge3 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| HLEHumanity's Last Exam | Score36.5% | Versus best verified row Best verified: Kimi K3.1 Pro · 70% | Gap33.5 behind | Weight44% ref. weight | Provider exact |
| GPQAGraduate-Level Google-Proof Q&A | Score91% | Versus best verified row Best verified: GPT-6 Astra · 96% | Gap5 behind | Weight3% ref. weight | Provider exact |
| GPQA-DGPQA Diamond | Score91.0% | Versus best verified row Best verified: GPT-6 Astra · 96.0% | Gap5 behind | WeightDisplay only | Provider exact |
Inst. Following1 row
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| IFBenchInstruction Following Benchmark | Score81.5% | Versus best verified row Best verified: Flint · 89.2% | Gap7.7 behind | WeightWeighted 70% | Provider exact |
Bars run 0–100; the dark tick marks the best source-verified value
All 19 rowsLineage
The sequence follows explicit supersedes links. Each score is estimated for that model; a relative can inform a sparse estimate but never sets a floor, so a newer release can score below an earlier one. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.
Base entry
Spec sheet
Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.
- API model ID
- qwen3.8-omni-flashAlibaba Cloud Model Studio pricing
- Context window
- 1MAlibaba Cloud Model Studio pricing
- Maximum output
- Not sourced yet
- Knowledge cutoff
- Not sourced yet
- Input modalities
- Not sourced yet
- Output modalities
- Not sourced yet
- Parameters
- Not disclosed by the provider
- Availability
- Hosted only. Qwen serves qwen3.8-omni-flash through the Qianwen AI Platform and Alibaba Cloud Model Studio with OpenAI-compatible Chat Completions and Responses APIs, text, image, audio and video input, a 1,000,000-token context window, and reasoning_effort levels of low, medium, and xhigh (default). A separate streaming sibling, qwen3.8-omni-flash-realtime, is served over WebSocket and WebRTC for continuous audio-visual interaction; its throughput and latency figures are not attributed to this row. No weights are published.
- Cloud regions
- Not tracked yet
- Lifecycle
- Current
- API capabilities
- Tool calling, structured outputs, and batch support are not tracked yet
- Prompt caching
- Not documented in the pricing recordAlibaba Cloud Model Studio pricing
- Self-host
- Weights are not published
- Rate limits
- Not tracked yet
How to read this profile
The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.
We track Qwen3.8-Omni-Flash, but the public leaderboard excludes this profile until enough non-generated benchmark coverage is available. Only published rows appear above.
Qwen3.8-Omni-Flash is a proprietary model with a 1M context window. It uses an explicit reasoning mode, which can improve complex problem solving while adding latency and token use.
Hosted only. Qwen serves qwen3.8-omni-flash through the Qianwen AI Platform and Alibaba Cloud Model Studio with OpenAI-compatible Chat Completions and Responses APIs, text, image, audio and video input, a 1,000,000-token context window, and reasoning_effort levels of low, medium, and xhigh (default). A separate streaming sibling, qwen3.8-omni-flash-realtime, is served over WebSocket and WebRTC for continuous audio-visual interaction; its throughput and latency figures are not attributed to this row. No weights are published.
Official exact-value snapshot from Qwen's September 18, 2026 Qwen3.8-Omni-Flash launch post. BenchLM maps only the text and vision table rows that match existing protocols. GPQA Diamond lands on both the weighted GPQA lane and the display GPQA-D lane, matching the existing provider-row convention; MathVision and CharXiv (RQ) keep their paired no-Code-Interpreter and Code-Interpreter values on separate keys; LiveCodeBench v6, DeepSWE 1.1, NL2Repo-Bench, and the in-house CoWorkBench stay display-only. HLE is stored on the weighted lane only, because Qwen labels the row as GPT-4o-judged without stating a no-tools setting. ClawEval-MM is reported as a pass@3 and average pair and stays out of schema, as on the Qwen3.8-Flash-Next row. The launch's 34-row omni table (WildClawBench-MM, UniClawBench, AgenticVBench, OmniGAIA, DailyOmni, WorldSense, AVUT, JoinAVBench, OmniVideoBench, Video-MME-v2, LVOmniBench, OmniCloze, OmniCap-IF, QIVD, StreamingBench, the four multi-speaker ASR sets, WenetSpeech, FLEURS ASR and S2TT, SpotSoundBench, MMAU, MMAR, MMSU, LongAudioSpan, the three music sets, Audio MultiChallenge, WildSpeech, and VoiceBench) has no equivalent in the weighted schema and is recorded as provider evidence on the voice-benchmark page instead of being collapsed into non-equivalent keys. The Agentic Omni Understanding table, which pairs static and Qwen Code agent runs on OmniVideoBench (63.4 to 67.8), Video-MME-v2 (65.0 to 71.3), and LVOmniBench (63.3 to 73.6), is a harness comparison and is recorded there too. The realtime throughput and latency table belongs to qwen3.8-omni-flash-realtime and is not stored on this row.
19 of 782 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.
Its strongest eligible category is Instruction Following at #23. a well-rounded choice across a range of tasks.
Last updated October 29, 2026. Runtime fields remain blank until a sourced snapshot exists.
Deployment options
Self-host and provider-specific paths stay separate from benchmark evidence so operating constraints are visible before a score becomes the whole decision.
Where to run Qwen3.8-Omni-Flash by API
OpenRouter is the one host we track that publishes a per-token price for it. Prices read from each host’s own pages on October 20, 2026.
| Host | Input /1M | Output /1M | Cached /1M | Context |
|---|---|---|---|---|
| OpenRouter | $0.15 | $0.47 | $0.016 | 1M |
Questions
How does Qwen3.8-Omni-Flash perform overall in AI benchmarks?
Qwen3.8-Omni-Flash has 19 source-displayable benchmark rows, but it does not qualify for a public overall rank. The available rows remain visible by category without being converted into a site-wide score. Missing evidence stays blank instead of being estimated from an earlier model.
Is Qwen3.8-Omni-Flash good for knowledge and understanding?
Qwen3.8-Omni-Flash has source-displayable benchmark coverage for knowledge and understanding, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is Qwen3.8-Omni-Flash good for coding and programming?
Qwen3.8-Omni-Flash has source-displayable benchmark coverage for coding and programming, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is Qwen3.8-Omni-Flash good for agentic tool use and computer tasks?
Qwen3.8-Omni-Flash has source-displayable benchmark coverage for agentic tool use and computer tasks, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is Qwen3.8-Omni-Flash good for multimodal and grounded tasks?
Qwen3.8-Omni-Flash has source-displayable benchmark coverage for multimodal and grounded tasks, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is Qwen3.8-Omni-Flash good for instruction following?
Qwen3.8-Omni-Flash ranks #23 out of 157 eligible models for instruction following, with a public category score of 91.6/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Does Qwen3.8-Omni-Flash have full benchmark coverage on BenchLM?
No. Qwen3.8-Omni-Flash currently has 19 source-displayable rows. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.
What is the context window size of Qwen3.8-Omni-Flash?
Qwen3.8-Omni-Flash has a documented context window of 1M. That figure is the maximum combined prompt and retained-conversation space reported for this exact model; it is not the maximum output length. The profile keeps output limits separate because providers often publish those limits independently.