Meituan model product
LongCat 2.5
5 is Meituan’s multimodal reasoning model for coding and agent tasks, with image understanding and a 1M-token context window.
Updated Nov 11, 2026. Default version: LongCat-2.5
LLMBoard Score81.0LongCat-2.5
Coverage20%19 benchmark families
Context window1MTokens
Official input priceN/AOfficial price unavailable
LongCat 2.5 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
LongCat 2.5 Benchmark Results
Benchmark scores for LongCat-2.5.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkFinSearchComp-T3 | Score63.40% | Rank01 | Participants2 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkFrontier Science | Score77.00% | Rank01 | Participants2 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkGDPval-Rubrics | Score84.80% | Rank01 | Participants2 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkSpreadsheetBench 2 | Score34.90% | Rank01 | Participants3 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkSWE Atlas - Test Writing | Score47.80% | Rank01 | Participants2 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkVision2Web | Score78.60% | Rank01 | Participants6 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkClaw-Eval | Score78.30% | Rank02 | Participants16 | Percentile93.33% | EvidenceC | Evaluated |
| BenchmarkHumanity's Last Exam (with tools, text-only) | Score63.10% | Rank02 | Participants12 | Percentile90.91% | EvidenceC | Evaluated |
| BenchmarkMathVision | Score96.50% | Rank02 | Participants40 | Percentile97.44% | EvidenceC | Evaluated |
| BenchmarkPaperBench | Score82.10% | Rank02 | Participants4 | Percentile66.67% | EvidenceC | Evaluated |
| BenchmarkBrowseComp | Score91.90% | Rank03 | Participants76 | Percentile97.33% | EvidenceC | Evaluated |
| BenchmarkSWE Atlas - Codebase QnA | Score58.30% | Rank03 | Participants9 | Percentile75.00% | EvidenceC | Evaluated |
| BenchmarkSkillsBench | Score63.30% | Rank04 | Participants13 | Percentile75.00% | EvidenceC | Evaluated |
| BenchmarkAutomationBench v1.0.6 | Score51.80% | Rank04 | Participants7 | Percentile50.00% | EvidenceC | Evaluated |
| BenchmarkNL2Repo | Score60.50% | Rank05 | Participants26 | Percentile84.00% | EvidenceC | Evaluated |
| BenchmarkJobBench | Score65.40% | Rank05 | Participants7 | Percentile33.33% | EvidenceC | Evaluated |
| BenchmarkCharXiv-R | Score90.20% | Rank07 | Participants64 | Percentile90.48% | EvidenceC | Evaluated |
| BenchmarkTerminal-Bench 2.1 | Score88.40% | Rank07 | Participants56 | Percentile89.09% | EvidenceC | Evaluated |
| BenchmarkVideo-MME | Score85.20% | Rank07 | Participants18 | Percentile64.71% | EvidenceC | Evaluated |
| BenchmarkMCP Atlas | Score83.10% | Rank08 | Participants42 | Percentile82.93% | EvidenceC | Evaluated |
| BenchmarkProgram Bench | Score25.50% | Rank17 | Participants20 | Percentile15.79% | EvidenceC | Evaluated |
| BenchmarkTerminal-Bench 4.0 | Score38.50% | Rank20 | Participants35 | Percentile44.12% | EvidenceC | Evaluated |
| BenchmarkDeepSWE 1.1 | Score64.60% | Rank33 | Participants54 | Percentile39.62% | EvidenceC | Evaluated |
LongCat 2.5 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 2.5 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- N/A
- Official provider
- N/A
- Lowest third-party
- From $0.30 input, $1.2 output per 1M via Vercel AI Gateway
- Tracked offerings
- 1
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderVercel AI Gateway | Provider model IDmeituan/longcat-2.5 | Regionglobal | Input / 1M$0.30 | Output / 1M$1.2 | Context1M | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
LongCat 2.5 Runtime Performance
Provider-specific output speed and catalog latency for LongCat-2.5. Runtime does not affect the capability score.
No runtime data
No provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
LongCat 2.5 Specifications
Technical details for the model's default version.
- Version
- LongCat-2.5
- Released
- Nov 1, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 1.6T
- Context window
- 1M
- Max output
- 131.1K
- Inputs
- image, text, video
- Outputs
- text
- Open weights
- No
- License
- Unknown
LongCat 2.5 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-2.5 | Released | LLMBoard80.96 | Parameters1.6T | Context1M | Max output131.1K | Open weightsNo | LicenseUnknown |
Models similar to LongCat 2.5
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
ME
LongCat Flash Thinking
Meituan
ME
LongCat Flash Chat
Meituan
ME
LongCat Flash Lite
Meituan
DE
DeepSeek-V4-Pro
DeepSeek
IN
Ling 3.1 Flash
Inclusionai
GO
Google
Gemini 3.8 Flash
What is LongCat 2.5?
Key information about LongCat 2.5 and its available data.
5 is a multimodal reasoning model from Meituan for coding and agent tasks. It supports image understanding and has a 1M-token context window.
Data as of 2026-11-10.
FAQ
Common questions about LongCat 2.5.
When was LongCat 2.5 released?
LongCat 2.5's default version was released on Nov 1, 2026.
How much does LongCat 2.5 cost?
No official standard PAYG price is currently available for LongCat 2.5. The lowest tracked third-party offer starts at $0.30 input and $1.2 output via Vercel AI Gateway.
Who created LongCat 2.5?
LongCat 2.5 was created by Meituan.
What is the context window for LongCat 2.5?
The default version has a 1M token context window.
Is LongCat 2.5 open weight?
No. The default version is not marked as having publicly available weights.
How many API providers offer LongCat 2.5?
1 provider offerings are linked to the default version.