Claude Opus 5 vs Gemini 3.1 Pro Preview
Claude Opus 5 versus Gemini 3.1 Pro Preview measures how far the frontier moved in five months: Opus leads the Intelligence Index 63 to 48. Google's Preview — the index leader at its February 2026 debut — still offers full multimodal input and fast, concise serving at $2/$12 per million tokens versus Opus's $5/$25.
Side by side
| Claude Opus 5 | Gemini 3.1 Pro Preview | |
|---|---|---|
| Maker | Anthropic | |
| Released | 2026-07-24 | 2026-02-19 |
| Context window | 1M tokens | 1M tokens |
| Modalities in | text, image | text, image, audio, video |
| Input $/MTok | $5 | $2 |
| Output $/MTok | $25 | $12 |
| Open weights | No | No |
Benchmark deltas
| Benchmark | Claude Opus 5 | Gemini 3.1 Pro Preview | Δ |
|---|---|---|---|
| AA Intelligence Index | 63 | 48 | +15 |
Δ is Claude Opus 5 minus Gemini 3.1 Pro Preview, in points. Shaded cell wins the row. Dashed values are vendor-published, not yet reproduced — methodology.
Cost calculator
- Claude Opus 5
- $500/mo
- Gemini 3.1 Pro Preview
- $220/mo
- Difference
- $280/mo
List prices as of 2026-08-11; excludes caching, batch, and long-context tiers — see model pages for tier notes.
Verdict
In February, Gemini 3.1 Pro Preview led the index; by August it trails the leader by fifteen points without shipping a successor — Gemini 3.5 Pro remains announced but unreleased. That is the sharpest illustration of 2026's release velocity in our dataset. The Preview remains a rational choice for multimodal document work at mid-tier prices, where its audio and video input beat Opus's text-and-image ceiling. For reasoning, coding, and agentic work, the gap is decisive. Google's next release resets this page; until then, Opus by a wide margin.
Verdicts are analysis, not measurement, and are labeled as such per our editorial policy.
FAQ
- Is Claude Opus 5 cheaper than Gemini 3.1 Pro Preview?
- No. Gemini 3.1 Pro Preview is cheaper: $2 input / $12 output per million tokens against $5 / $25 for Claude Opus 5 (prices as of August 11, 2026).
- Which has the longer context window — Claude Opus 5 or Gemini 3.1 Pro Preview?
- They match: both accept 1M tokens of context.
- Which is the more capable model, Claude Opus 5 or Gemini 3.1 Pro Preview?
- Claude Opus 5 scores higher on the Artificial Analysis Intelligence Index: 63 (measured August 11, 2026) against 48 for Gemini 3.1 Pro Preview (measured August 11, 2026).
Related news
Comparison published August 11, 2026; underlying data last changed August 11, 2026.