DeepSeek V3
DeepSeek V3Alternatives?
CheaperGLM 4.7 FlashthinkingOpen×4.3-54%$0.061 / $0.40
BetterMinistral 3 3B 2512Open+9%-60%$0.008OpenGLM 5.2max×4.3+6,396%$1.28Memory (4-bit): 411 GB → 452 GB
CheaperHy3 previewthinkingOpen×4.1-10%$0.18 / $0.60No strict JSON
CheaperQwen3.5 4BthinkingOpen×4-81%$0.03 / $0.15GLM 4.7 FlashthinkingOpen×4.3-54%$0.061 / $0.40Hy3 previewthinkingOpen×4.1-10%$0.18 / $0.60No strict JSONQwen3.5 4BthinkingOpen×4-81%$0.03 / $0.15
MiMo v2 Omnithinking×4-44%$0.14 / $0.28Qwen 3.5 9Bnon-thinkingOpen×3.7-71%$0.08 / $0.13Ministral 3 3B 2512Open+9%-60%$0.008
GPT OSS 20BhighOpen×2.6-47%$0.011Ministral 3 8B 2512Open+17%-44%$0.011
Nemotron 3 Nano 30B A3BthinkingOpen+79%-15%$0.017Mistral Small 4thinkingOpen+81%-13%$0.017GLM 5.2maxOpen×4.3+6,396%$1.28Memory (4-bit): 411 GB → 452 GBGLM 4.7 FlashthinkingOpen×4.3-54%$0.061 / $0.40
Step 3.7 FlashhighOpen×4.3+39%$0.20 / $1.15GLM 5thinkingOpen×4.3+392%$1 / $3.20Memory (4-bit): 411 GB → 452 GBGLM 5.1thinkingOpen×4.3+3,691%$0.75Memory (4-bit): 411 GB → 452 GBDeepSeek V3 0324non-thinkingOpen×2.1+4%$0.021DeepSeek V3.1 Terminusnon-thinkingOpen+63%—$0.27 / $0.95R1 0528thinkingOpen+60%—$0.39 / $1.63DeepSeek V3.1non-thinkingOpen+53%—$0.25 / $0.95DeepSeek V3.2 SpecialethinkingOpen-100%——DeepSeek R1 0528 Qwen3 8BthinkingOpen-100%—$0.06 / $0.09Context: 128K → 33KDeepSeek V4.1 Flash 1017maxOpenn/a—$0.26Memory (4-bit): 411 GB → 458 GB+5DeepSeek V4 Pro 0813non-thinkingOpenn/a—$0.36Memory (4-bit): 411 GB → 990 GB+5All DeepSeek models →Run it yourself
estimateHistory
5 versionsBenchmarks
in Agents & Tool UseFAQ
How much does DeepSeek V3 cost?›
$0.28 per 1M input tokens and $0.42 per 1M output tokens; cached input $0.028 per 1M. A typical benchmark task costs about $0.02 at default effort.
Is DeepSeek V3 free to use?›
There is no free API; the cheapest paid price is $0.28 per 1M input tokens. The weights are open, so you can run it yourself.
How good is DeepSeek V3, and what is it best at?›
#331 of 336 in Chat & Knowledge, #77 of 134 in Coding, #119 of 135 in Agents & Tool Use, #113 of 197 in Writing & Language, #124 of 135 in Vision & Documents.
Is DeepSeek V3 better than Magistral Medium 1?›
On τ²-bench · AA, DeepSeek V3 scores 22.8 vs 23.1 — about the same.
How can I use DeepSeek V3?›
Through the API of DeepInfra, Novita.
What is the context window of DeepSeek V3?›
128K tokens, with up to 8K output tokens.
When was DeepSeek V3 released? Is there a newer version?›
Released 2024-12-26. It has been superseded by DeepSeek V3.1.
Is DeepSeek V3 open source?›
Yes, the weights are open.
Can I run DeepSeek V3 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can DeepSeek V3 read images?›
No, it accepts text only.
What is DeepSeek V3?›
DeepSeek V3 is an open-weights language model from DeepSeek, released Dec 2024.
What are the best alternatives to DeepSeek V3?›
GLM 4.7 Flash (cheaper), Ministral 3 3B 2512 (better), GLM 5.2 (open weights).