DeepSeek model product
DeepSeek-V2.5
5 is an upgraded large language model that combines general-purpose and coding capabilities.
Updated Oct 23, 2026. Default version: DeepSeek-V2.5
LLMBoard Score12.7DeepSeek-V2.5
Coverage20%11 benchmark families
Context window8.2KTokens
Official input priceN/AOfficial price unavailable
DeepSeek-V2.5 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
DeepSeek-V2.5 Benchmark Results
Benchmark scores for DeepSeek-V2.5.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkAider | Score72.20% | Rank01 | Participants4 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkDS-Arena-Code | Score63.10% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkDS-FIM-Eval | Score78.30% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkLiveCodeBench(01-09) | Score41.80% | Rank01 | Participants1 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkHumanEval-Mul | Score73.80% | Rank02 | Participants2 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkAlignBench | Score80.40% | Rank03 | Participants5 | Percentile50.00% | EvidenceC | Evaluated |
| BenchmarkAlpacaEval 2.0 | Score50.50% | Rank03 | Participants4 | Percentile33.33% | EvidenceC | Evaluated |
| BenchmarkMT-Bench | Score90.20% | Rank03 | Participants13 | Percentile83.33% | EvidenceC | Evaluated |
| BenchmarkBBH | Score84.30% | Rank05 | Participants13 | Percentile66.67% | EvidenceC | Evaluated |
| BenchmarkArena Hard | Score76.20% | Rank09 | Participants27 | Percentile69.23% | EvidenceC | Evaluated |
| BenchmarkGSM8k | Score95.10% | Rank11 | Participants48 | Percentile78.72% | EvidenceC | Evaluated |
| BenchmarkHumanEval | Score89.00% | Rank16 | Participants66 | Percentile76.92% | EvidenceC | Evaluated |
| BenchmarkMATH | Score74.70% | Rank29 | Participants71 | Percentile60.00% | EvidenceC | Evaluated |
| BenchmarkMMLU | Score80.40% | Rank62 | Participants103 | Percentile40.20% | EvidenceC | Evaluated |
| BenchmarkSWE-Bench Verified | Score16.80% | Rank119 | Participants120 | Percentile0.84% | EvidenceC | Evaluated |
| BenchmarkLM Arena Text | Score1,293.95 rating | Rank172 | Participants218 | Percentile21.20% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text Style Control | Score1,323.42 rating | Rank175 | Participants218 | Percentile19.82% | EvidenceA | Evaluated |
DeepSeek-V2.5 Arena Results
Preference and agent-evaluation results for the default version.
Sort by
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenatext | Categoryjapanese | Rank147 | Rating / score1,227.92 | Votes161 | ObservationsN/A | Result date |
| Arenatext style control | Categoryjapanese | Rank148 | Rating / score1,252.03 | Votes161 | ObservationsN/A | Result date |
| Arenatext | Categorykorean | Rank151 | Rating / score1,209.34 | Votes344 | ObservationsN/A | Result date |
| Arenatext style control | Categorykorean | Rank155 | Rating / score1,254.73 | Votes344 | ObservationsN/A | Result date |
| Arenatext | Categoryfrench | Rank156 | Rating / score1,289.37 | Votes222 | ObservationsN/A | Result date |
| Arenatext style control | Categorycreative writing | Rank157 | Rating / score1,309.44 | Votes1,090 | ObservationsN/A | Result date |
| Arenatext style control | Categorychinese | Rank161 | Rating / score1,355.31 | Votes446 | ObservationsN/A | Result date |
| Arenatext style control | Categoryfrench | Rank161 | Rating / score1,326.28 | Votes222 | ObservationsN/A | Result date |
| Arenatext | Categorychinese | Rank162 | Rating / score1,317.67 | Votes446 | ObservationsN/A | Result date |
| Arenatext | Categorygerman | Rank162 | Rating / score1,258.20 | Votes140 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry writing and literature and language | Rank162 | Rating / score1,314.86 | Votes1,930 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry entertainment and sports and media | Rank164 | Rating / score1,302.10 | Votes1,216 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry software and it services | Rank165 | Rating / score1,364.36 | Votes1,688 | ObservationsN/A | Result date |
| Arenatext style control | Categorylonger query | Rank165 | Rating / score1,337.33 | Votes1,037 | ObservationsN/A | Result date |
| Arenatext style control | Categoryrussian | Rank165 | Rating / score1,323.30 | Votes1,034 | ObservationsN/A | Result date |
| Arenatext | Categorylonger query | Rank166 | Rating / score1,300.59 | Votes1,037 | ObservationsN/A | Result date |
| Arenatext | Categoryrussian | Rank166 | Rating / score1,289.21 | Votes1,034 | ObservationsN/A | Result date |
| Arenatext style control | Categorycoding | Rank166 | Rating / score1,373.74 | Votes1,079 | ObservationsN/A | Result date |
| Arenatext | Categorycreative writing | Rank167 | Rating / score1,284.89 | Votes1,090 | ObservationsN/A | Result date |
| Arenatext | Categoryexpert | Rank167 | Rating / score1,266.14 | Votes441 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry entertainment and sports and media | Rank167 | Rating / score1,275.03 | Votes1,216 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry software and it services | Rank167 | Rating / score1,313.24 | Votes1,688 | ObservationsN/A | Result date |
| Arenatext style control | Categorygerman | Rank167 | Rating / score1,290.02 | Votes140 | ObservationsN/A | Result date |
| Arenatext | Categorycoding | Rank168 | Rating / score1,309.28 | Votes1,079 | ObservationsN/A | Result date |
| Arenatext | Categorymath | Rank168 | Rating / score1,288.48 | Votes1,031 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts english | Rank169 | Rating / score1,298.94 | Votes1,132 | ObservationsN/A | Result date |
| Arenatext | Categoryinstruction following | Rank169 | Rating / score1,279.55 | Votes2,970 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts | Rank170 | Rating / score1,289.19 | Votes1,790 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry medicine and healthcare | Rank171 | Rating / score1,292.00 | Votes291 | ObservationsN/A | Result date |
| Arenatext | Categorymulti turn | Rank171 | Rating / score1,296.65 | Votes1,178 | ObservationsN/A | Result date |
DeepSeek-V2.5 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.
DeepSeek-V2.5 Runtime Performance
Provider-specific output speed and catalog latency for DeepSeek-V2.5. Runtime does not affect the capability score.
Sort by
Provider | Output Speed | Catalog Latency | Max Input | Max Output | Updated |
|---|
| ProviderDeepSeek | Output Speed100.00 tok/s | Catalog Latency0.50 s | Max Input8.2K | Max Output8.2K | Updated |
| ProviderDeepInfra | Output Speed63.00 tok/s | Catalog Latency0.50 s | Max Input8.2K | Max Output8.2K | Updated |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
DeepSeek-V2.5 Specifications
Technical details for the model's default version.
- Version
- DeepSeek-V2.5
- Released
- May 8, 2024
- Knowledge cutoff
- Unknown
- Parameters
- 236B
- Context window
- 8.2K
- Max output
- 8.2K
- Inputs
- text
- Outputs
- text
- Open weights
- Yes
- License
- deepseek
DeepSeek-V2.5 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 |
|---|
| VersionDeepSeek-V2.5 | Released | LLMBoard12.69 | Parameters236B | Context8.2K | Max output8.2K | Open weightsYes | Licensedeepseek |
Models similar to DeepSeek-V2.5
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
DE
DeepSeek-R1-Distill-Qwen
DeepSeek
DE
DeepSeek-R1-Distill-Llama
DeepSeek
DE
DeepSeek-V3
DeepSeek
DE
DeepSeek-VL2
DeepSeek
DE
DeepSeek-V3.1
DeepSeek
DE
DeepSeek-R1
DeepSeek
What is DeepSeek-V2.5?
Key information about DeepSeek-V2.5 and its available data.
5 combines DeepSeek-V2-Chat and DeepSeek-Coder-V2-Instruct, integrating general and coding abilities. It is optimized for writing, instruction following, and alignment with human preferences.
Data as of 2026-10-22.
FAQ
Common questions about DeepSeek-V2.5.
When was DeepSeek-V2.5 released?
DeepSeek-V2.5's default version was released on May 8, 2024.
How much does DeepSeek-V2.5 cost?
No official standard PAYG price is currently available for DeepSeek-V2.5.
Who created DeepSeek-V2.5?
DeepSeek-V2.5 was created by DeepSeek.
What is the context window for DeepSeek-V2.5?
The default version has a 8.2K token context window.
Is DeepSeek-V2.5 open weight?
Yes. The default version is marked as open weight under deepseek.
How many API providers offer DeepSeek-V2.5?
No provider offering is currently linked to the default version.