Open-source 2B decision model from AWS's Strands Agents project that picks options and returns scores with a reliability figure
DeepSeek V4 Pro
New· 4DeepSeek's production flagship with a 1M-token context, built for tool-using agent work
About DeepSeek V4 Pro
DeepSeek released the production version of its flagship, V4 Pro, on 12 August 2026. The release, labelled build 0813, closed out a preview period of close to four months.
Context, output and thinking modes
V4 Pro works with 1 million tokens of context and can write as many as 384,000 tokens in one response. A switch lets it run in thinking or non-thinking mode.
Built for agent work
The general-availability release is aimed squarely at agent use: tool calling, code execution and multi-step workflows. Reported results are 87.9 for Terminal Bench 2.1, 62.7 for DeepSWE and 61.5 for NL2Repo.
A pricing change that arrived with it
From 16 August 2026 DeepSeek moved to peak and off-peak API rates, with tokens in peak hours costing double those off-peak. Buyers who can schedule work outside peak hours may benefit from the new structure.
TrustList summary
DeepSeek V4 Pro is DeepSeek's flagship model, released in production form on 12 August 2026 as build 0813. The release ended a preview that had run for nearly four months. It handles a context of 1M tokens and can produce up to 384,000 tokens in one response, with a switch between thinking and non-thinking modes.
Main features
- Tool-using agents: aimed at tool use, code execution and multi-step workflows.
- Thinking modes: a switch between thinking and non-thinking operation.
- Long output: up to 384,000 tokens in a single response.
- Benchmark scores: Terminal Bench 2.1 at 87.9, DeepSWE at 62.7 and NL2Repo at 61.5.
Key facts
- Released: 12 August 2026, ending a preview of nearly four months
- Context window: 1M tokens
- Build: 0813
Summary written by TrustList from the business's own published information, Nov 2026.
Benchmarks & AI stats
| Benchmark | Official | Community avg |
|---|---|---|
| Artificial Analysis Intelligence Index v5.0.1llm | 23 index | - |
| DeepSWEcode | 62.7 % | - |
| NL2Repocode | 61.5 % | - |
| Terminal Bench 2.1agent | 87.9 % | - |
“Official” values are editor-approved and feed the ranking. “Community avg” is the mean of member submissions (shown for transparency; it never affects the ranking until an editor approves a value).
Sign in to add a benchmark score for this model.
DeepSeek V4 Pro launches
Trust Score
An earned signal from verification, reviews, awards, transparency and engagement - the vendor can't buy it.
Updated 9/16/2026
What the business tells TrustList
- Founded
- 2026 (0 years) evidence sent
Request a demo or quote from DeepSeek V4 Pro
Reviews
Write the first review of DeepSeek V4 Pro
Used it? Your experience helps other buyers decide.
Write a reviewQuestions & answers
No questions yet. Be the first to ask about DeepSeek V4 Pro.
DeepSeek V4 Pro alternatives
Similar listings buyers compare against this one.
Perplexity's open-weight decision model that picks among given options for routing and checks in AI systems
Microsoft AI's text-to-speech model for keeping one voice across 23 languages, with voice cloning from short audio
Liquid AI's open-weight decision model, giving yes or no, choice or score answers from text and images
Microsoft AI's real-time speech-to-text model for voice agents, transcribing 60 languages with continuous language detection
Open-weight translation model from Bilibili's Index team, covering 150 languages and following glossary and style instructions
Cantina Security's first open-weight security model, a post-trained GLM-5.3-Flash for vulnerability research and exploit checks
Aleph Alpha's open-weight English-German mixture-of-experts model for self-hosting, released under Apache 2.0 in October 2026
More on TrustList
Everything here links back to the same verified catalogue. Pick your next stop.
- Write a reviewHelp other buyers decide
- CompaniesAgencies, consultancies and IT service providers, ranked by verified reviews.
- ProductsSoftware and SaaS with pricing, features, integrations and alternatives.
- ArticlesGuides, comparisons and research from the editorial desk and the community.
- AwardsAnnual recognition decided by verified reviews and an independent jury.
- LaunchesNew products and releases, voted up by the community every day.
- RequestsBuyers describe what they need; vendors respond directly.
- PeopleReviewers, authors and makers with public profiles.
- ComparePut up to four listings side by side before you shortlist.