Gemini 3 Pro
Google's flagship multimodal model with a 1M-token context window and a dedicated Deep Think reasoning mode.
About
Gemini 3 Pro is Google's flagship foundation model, announced November 18, 2025 as the successor to Gemini 2.5 Pro. It reads and generates across text, image, audio, and video with a one-million-token input context window and up to 64,000 tokens of output, and it shipped day one across the Gemini app, Google AI Studio, Vertex AI, and Google Search. An optional Deep Think mode pushes reasoning further for hard math and science problems.
Google prices the API at $2.00 per million input tokens and $12.00 per million output tokens for prompts up to 200K tokens, doubling beyond that, with a free rate-limited tier in AI Studio. At launch it broke 1500 Elo on the LMArena text leaderboard (1501, third-party measurement) and Google reported 81% on MMMU-Pro and 41% on Humanity's Last Exam in Deep Think mode. Google followed it with Gemini 3.1 Pro in February 2026, which kept the same context window and pricing while pushing reasoning scores higher.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
- LMArena Text Arena: 1501 Elo, #1 at launch (third-party measurement, per best-of-ai)
- MMMU-Pro: 81% multimodal reasoning (Google-reported, via max-productive.ai comparison)
- SWE-bench Verified: 76.2% (third-party comparison; Claude Sonnet 4.5 reported 77.2% on the same table)
- Humanity's Last Exam: 41% in Deep Think mode vs 37.5% standard (Google-reported)
- GPQA Diamond: 93.8% in Deep Think mode (Google-reported)
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 9/10
Scored 9/10 (verified tier, metrics-v1). Evidence: Google's flagship since Nov 18, 2025 with day-one rollout across the Gemini app, AI Studio, Vertex AI, and Search; 1M-token context with published $2/$12 per 1M pricing; LMArena #1 at launch (1501 Elo, third-party). Closed weights. Launch facts corroborated across Google docs and multiple independent outlets. Strengths: native multimodality, strong reasoning scores, aggressive pricing for its class. Docked for: (1) headline benchmark figures (MMMU-Pro, HLE, GPQA) are vendor-reported without independent replication; (2) closed weights with no training-data transparency.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 10.0/10
- Maintenance · recency, releases, issue load
- 10.0/10
- Openness · license + docs
- 2.0/10
- Manual review · sanity read + red-flag check
- 9.0/10
- · Google flagship since 2025-11-18; day-one rollout across Gemini app, AI Studio, Vertex AI, Search
- · LMArena 1501 Elo #1 at launch (third-party, best-of-ai); published $2/$12 per 1M pricing
- · Gemini 3.1 Pro followed Feb 2026 — active maintenance
- · Headline benchmarks (MMMU-Pro 81%, HLE 41%) are Google-reported; closed weights