DeepSeek Coder v2
Jun 2024
DeepSeek Coder v2Compare inon
Benchmarks of DeepSeek Coder v214
BenchmarkByScoreRank
No benchmark matches
—per task?
price not verifiedprice
—output speed?
—context
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
BetterGPT-5 Minihighwins 66%—$0.053
OpenGLM 5.3 Flashunstatedwins 79%—$0.15 / $0.50Memory (4-bit): 141 GB → 193 GB+5BetterGPT-6 Lunamaxwins 76%—$0.068Function calling only with none · Chat Completions+3BetterGPT-5.6 Lunaxhighwins 76%—$0.085BetterGPT-5.6 Terraxhighwins 78%—$0.63Prompts with >272K input tokensModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT-5 Minihighwins 66%—$0.053GPT-6 Lunamaxwins 76%—$0.068Function calling only with none · Chat Completions+3GPT-5.6 Lunaxhighwins 76%—$0.085GPT-5.6 Terraxhighwins 78%—$0.63Prompts with >272K input tokensGPT-6.1 Solmaxwins 81%—$0.72Remove unsupported parameters+5ModelScoreCostPer taskWhat changes
GLM 5.3 FlashunstatedOpenwins 79%—$0.15 / $0.50Memory (4-bit): 141 GB → 193 GB+5
Hy3unstatedOpenwins 78%—$0.13 / $0.53Memory (4-bit): 141 GB → 179 GB
Gemma 4 31BunstatedOpenwins 77%—$0.10 / $0.34
Qwen3.5 397B A17BunstatedOpenwins 76%—$0.60 / $3.60Memory (4-bit): 141 GB → 242 GB
MiniMax M3unstatedOpenwins 76%—$0.30 / $1.20Memory (4-bit): 141 GB → 256 GB+2ModelScoreCostPer taskWhat changes
DeepSeek V4 Pro 0813highOpenwins 80%—$1.32 / $3.96Memory (4-bit): 141 GB → 990 GB+5DeepSeek V4.1 Flash 1017unstatedOpenwins 76%—$0.30 / $1.20Memory (4-bit): 141 GB → 458 GB+5DeepSeek V3.1 TerminusunstatedOpenwins 76%—$0.27 / $0.95Memory (4-bit): 141 GB → 411 GBDeepSeek V3.1thinkingOpenwins 76%—$0.25 / $0.95Memory (4-bit): 141 GB → 411 GBR1 0528unstatedOpenwins 74%—$0.39 / $1.63Memory (4-bit): 141 GB → 411 GBDeepSeek V3unstatedOpenwins 70%—$0.28 / $0.42Memory (4-bit): 141 GB → 411 GBDeepSeek v2.5unstatedOpenwins 65%——DeepSeek V3.2 SpecialethinkingOpenn/a——Memory (4-bit): 141 GB → 411 GBAll DeepSeek models →Run it yourself
estimateFits on✗ 8 GB (phone, laptop)✗ 16 GB✗ 24 GB GPU✗ 48 GB GPU✗ 80 GB GPU
Licencedeepseek-licensecommercial use with conditions
Benchmarks
all categoriesChat & Knowledge
Arena Text · overall1,265#172
Arena Text · Chinese1,200.8#161
Arena Text · English1,194.4#172
Arena Text · French1,184.8#136
Arena Text · German1,164.3#138
Arena Text · instruction following1,180.2#166
Arena Text · Japanese1,126.3#122
Arena Text · Korean1,104#132
Arena Text · non-English1,182#166
Arena Text · Russian1,187.8#165
Arena Text · Spanish1,153.1#146
Arena Text · overall · full cohort1,191.2#168
Writing & Language
Arena Text · creative writing1,205#175
Arena Text · creative writing · full cohort1,120.1#175
FAQ
How much does DeepSeek Coder v2 cost?›
There is no per-unit API price: price not verified.
How good is DeepSeek Coder v2, and what is it best at?›
#172 of 196 in Chat & Knowledge, #175 of 195 in Writing & Language.
Is DeepSeek Coder v2 better than Granite 4.2 3B?›
On Arena Text · creative writing, DeepSeek Coder v2 scores 1,205 vs 1,206 — about the same.
When was DeepSeek Coder v2 released? Is there a newer version?›
Released 2024-06-17. It is the current version in its line.
Is DeepSeek Coder v2 open source?›
Yes, the weights are open (deepseek-license licence).
Can I run DeepSeek Coder v2 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can DeepSeek Coder v2 read images?›
No, it accepts text only.
What is DeepSeek Coder v2?›
DeepSeek Coder v2 is an open-weights language model from DeepSeek, released Jun 2024.
What are the best alternatives to DeepSeek Coder v2?›
GPT-5 Mini (better), GLM 5.3 Flash (open weights).