DeepSeek V4 Pro
DeepSeek V4 Pro
DeepSeek's reasoning flagship, now out of preview and outscored by its own Flash model.
Updated
The read
V4 Pro left preview in August as V4-Pro-0813: a very large sparse model that thinks before it answers, with MIT-licensed weights and a million-token window. It is still good on hard reasoning, but V4.1 Flash, built on DeepSeek's newer architecture, now scores higher on both independent leaderboards at about a quarter of the price. Pro remains a capable MIT-licensed reasoning model, but for most new work Flash is the better buy.
Where it shines
- MIT licence, including for the reasoning traces
- Very long context, so whole repositories fit in one call
- Off-peak pricing halves the bill
Worth knowing first
- V4.1 Flash scores higher at about a quarter of the price
- Self-hosting the full model is a serious hardware commitment
Index score
83.4
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
- Rank
- 28 of 73
- Builds
- LLMs
- Category
- Reasoning
- Output
- MIT-licensed weights, plus a cheap API
- Pricing
- Free tier
- Best for
- Hard reasoning on MIT-licensed weights
Scores are our own editorial judgement, weighted craft 55, speed 10, control 15 and value 20.
Compare it with
Also worth a look
Full index- 21GPT-6 AstraGPT-6 AstraProvisionalOpenAI's most capable model: a specialist for the hardest problems, priced like one.Reasoning modelPaid only84.9
- 25GPT-5.6 SolGPT-5.6 SolProvisionalStrong and now discounted, but GPT-6.1 Sol beats it for half the money.Reasoning modelPaid only84.4
- 44Hunyuan HY3Hunyuan HY3ProvisionalSwitches between fast answers and slow thinking, and charges almost nothing for either.Reasoning modelPaid only80.0
