Meituan model product
LongCat Flash Thinking
LongCat-Flash-Thinking is Meituan's reasoning model built on the LongCat-Flash foundation with 560B total parameters (MoE, ~27B activated).
Updated Aug 17, 2026. Default version: LongCat-Flash-Thinking-2601
LLMBoard Score60.3LongCat-Flash-Thinking-2601
Coverage80%11 benchmark families
Context window128KTokens
Official input priceN/AOfficial price unavailable
LongCat Flash Thinking Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
LongCat Flash Thinking Benchmark Results
Benchmark scores for LongCat-Flash-Thinking-2601.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkTau2 Airline | Score76.5% | Rank01 | Participants23 | Percentile100.0% | EvidenceC | Evaluated |
| BenchmarkTau2 Telecom | Score99.3% | Rank02 | Participants35 | Percentile97.1% | EvidenceC | Evaluated |
| BenchmarkBrowseComp-zh | Score69.0% | Rank04 | Participants13 | Percentile75.0% | EvidenceC | Evaluated |
| BenchmarkTau2 Retail | Score88.6% | Rank04 | Participants26 | Percentile88.0% | EvidenceC | Evaluated |
| BenchmarkLiveCodeBench | Score82.8% | Rank08 | Participants75 | Percentile90.5% | EvidenceC | Evaluated |
| BenchmarkAIME 2025 | Score99.6% | Rank09 | Participants115 | Percentile93.0% | EvidenceC | Evaluated |
| BenchmarkIMO-AnswerBench | Score78.6% | Rank19 | Participants20 | Percentile5.3% | EvidenceC | Evaluated |
| BenchmarkBrowseComp | Score56.6% | Rank39 | Participants62 | Percentile37.7% | EvidenceC | Evaluated |
| BenchmarkHumanity's Last Exam | Score25.2% | Rank53 | Participants99 | Percentile46.9% | EvidenceC | Evaluated |
| BenchmarkSWE-Bench Verified | Score70.0% | Rank64 | Participants111 | Percentile42.7% | EvidenceC | Evaluated |
| BenchmarkGPQA | Score80.5% | Rank90 | Participants239 | Percentile62.6% | EvidenceC | Evaluated |
LongCat Flash Thinking Arena Results
Preference and agent-evaluation results for the default version.
No Arena results
The default version does not have a matching Arena result yet.
LongCat Flash Thinking Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- N/A
- Official provider
- N/A
- Lowest third-party
- N/A
- Tracked offerings
- 0
No provider prices
The default version has no current input or output token prices.
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
LongCat Flash Thinking Runtime Performance
Provider-specific output speed and catalog latency for LongCat-Flash-Thinking-2601. Runtime does not affect the capability score.
Sort by
Provider | Output Speed | Catalog Latency | Max Input | Max Output | Updated |
|---|
| ProviderMeituan | Output Speed100 tok/s | Catalog Latency3 s | Max Input128K | Max Output128K | Updated |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
LongCat Flash Thinking Specifications
Technical details for the model's default version.
- Version
- LongCat-Flash-Thinking-2601
- Released
- Jan 14, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 560B
- Context window
- 128K
- Max output
- 128K
- Inputs
- text
- Outputs
- text
- Open weights
- No
- License
- MIT
LongCat Flash Thinking Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionLongCat-Flash-Thinking-2601 | Released | LLMBoard60.3 | Parameters560B | Context128K | Max output128K | Open weightsNo | LicenseMIT |
| VersionLongCat-Flash-Thinking | Released | LLMBoardN/A | Parameters560B | Context128K | Max output128K | Open weightsNo | LicenseMIT |
LongCat Flash Thinking vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to LongCat Flash Thinking
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
ME
LongCat Flash Chat
Meituan
ME
LongCat Flash Lite
Meituan
ST
Step 3.5 Flash
StepFun
OP
GPT-5.4-mini
OpenAI
MA
Kimi K2 Thinking
Moonshot AI
AC
Qwen3.6 27B
Alibaba Cloud / Qwen Team
What is LongCat Flash Thinking?
Key information about LongCat Flash Thinking and its available data.
LongCat-Flash-Thinking is Meituan's reasoning model built on the LongCat-Flash foundation with 560B total parameters (MoE, ~27B activated). It introduces a training pipeline specifically tuned for advanced reasoning, featuring Re-thinking Mode that delivers parallel reasoning paths for sophisticated decision-making.
Achieves strong performance on mathematical reasoning, agentic tool use, and formal theorem proving benchmarks.
Data as of 2026-08-17.
FAQ
Common questions about LongCat Flash Thinking.
When was LongCat Flash Thinking released?
LongCat Flash Thinking's default version was released on Jan 14, 2026.
How much does LongCat Flash Thinking cost?
No official standard PAYG price is currently available for LongCat Flash Thinking.
Who created LongCat Flash Thinking?
LongCat Flash Thinking was created by Meituan.
What is the context window for LongCat Flash Thinking?
The default version has a 128K token context window.
Is LongCat Flash Thinking open weight?
No. The default version is not marked as having publicly available weights.
How many API providers offer LongCat Flash Thinking?
No provider offering is currently linked to the default version.
What models should I compare LongCat Flash Thinking with?
Nearby ranked alternatives include GPT-5.1-Instant, Kimi K2 Thinking, Step 3.5 Flash.