Ant GroupLing 3.0 Flash
Ling 3.0 Flash is an Ant Group open-weights reasoning model, released 4 Aug 2026. Its best configuration ranks #283 of 756 on the BenchLeader Index at 53.0 ±3.6, in the upper half. The ± is the point: 276 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in long context (62) and lowest in reasoning (44). At $0.115 per million tokens blended it is among the cheapest fifth of ranked models. Output speed of 325 tokens per second puts it in the fastest quarter, with a first answer in 9.0 s. Last measured 29 Sept 2026.
- Blended price
- $0.115/M
- $0.080 in · $0.220 out
- Output speed
- 325 tok/s
- measured by Artificial Analysis
- First answer
- 8.98 s
- first token 0.95 s
- Context
- 262k
- Full answer
- 11 s
- median, reasoning included
- Cost per run
- –
- Released
- 4 Aug 2026
- Overall index53
- Reasoning44
- Coding61
- Agents & tools50
- Knowledge53
- Long context62
- Composite53
Versions
Ant Group has shipped 3 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Ling 3.0 Flashthis page | 4 Aug 2026 | 53.0 | #283 |
| Ling Flash 2.0 | 18 Sept 2025 | 44.6 | #511 |
| NewLing 3.1 Flash | – | – | – |
How its index has moved
9 Sept 2026 to 2 Oct 2026The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| GPQA Diamond (AA)not in index | 85.5%#122 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 23.7%#173 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | 84.8%#48 | Vals AI | |
| CritPt | 1.7%#186 | Artificial Analysis |
Coding
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| SciCode (AA)not in index | 42.0%#145 | Artificial Analysis | |
| LiveCodeBench | 84.0%#42 | Vals AI | |
| SWE-bench (Vals)not in index | 65.2%#70 | Vals AI | |
| Code Migrationnot in index | 0.0%#67 | Vals AI | |
| Vibe Code Bench v1.1not in index | 2.9%#95 | Vals AI | |
| Terminal-Bench 4.0 (AA)not in index | 0.0%#137 | Artificial Analysis | |
| Terminal-Bench 2.1 (AA)not in index | 55.4%#103 | Artificial Analysis |
Agents & tools
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Terminal-Bench 2.1 (Vals) | 50.2%#57 | Vals AI | |
| GDPval (AA)not in index | 22.0%#153 | Artificial Analysis | |
| τ²-Bench Banking (AA) | 27.2%#74 | Artificial Analysis | |
| Finance Agent v2not in index | 30.3%#61 | Vals AI | |
| Harvey's Legal Agent Benchmarknot in index | 1.3%#45 | Vals AI | |
| Legal Research Benchnot in index | 0.0%#68 | Vals AI | |
| SkillsBenchnot in index | 27.9%#32 | Vals AI |
Knowledge
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-Omniscience | -17.9#219 | Artificial Analysis | |
| MMLU-Pro | 82.0%#76 | Vals AI | |
| LegalBench | 79.7%#89 | Vals AI | |
| CorpFin | 61.9%#50 | Vals AI | |
| TaxEval | 70.7%#85 | Vals AI | |
| Excel Modeling Benchmarknot in index | 25.0%#58 | Vals AI | |
| MedCodenot in index | 32.3%#84 | Vals AI | |
| MedScribenot in index | 80.9%#50 | Vals AI | |
| Public Benefits Benchnot in index | 50.7%#39 | Vals AI | |
| AA-Omniscience: accuracynot in index | 18.2%#322 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 56.0%#81 | Artificial Analysis |
Long context
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-LCR | 73.0%#168 | Artificial Analysis |
Composite
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA Intelligence Index | 20.1#216 | Artificial Analysis |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| NovitaAI | 128 tok/s | 0.89 s | $0.021 | $0.063 | 262k | – |
| DeepInfra | 52 tok/s | 0.51 s | $0.060 | $0.180 | 131k | bf16 |
Price history
Listed price per 1M tokens over time, as recorded by OpenRouter for the provider with the longest history.
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0001 | – | 9.9 s |
| Summarise a 30-page report | 12,000 / 600 | $0.0011 | – | 10.8 s |
| Code edit | 6,000 / 1,500 | $0.0008 | – | 13.6 s |
| Agentic coding session | 60,000 / 4,000 | $0.0057 | – | 21.3 s |
| Structured extraction | 2,000 / 200 | $0.0002 | – | 9.6 s |
See also
Data as of 5 Oct 2026. Compare with another model.
Cite as: BenchLeader, “Ling 3.0 Flash: benchmarks, pricing, speed and rank”, https://www.benchleader.com/models/ling-3-0-flash, data as of 5 Oct 2026.