Fast Gemini model balancing multimodal reasoning, tool use, and cost
Specification
google listing
Context
1M
Max output
66K
Released
2026-07-21
Knowledge cutoff
2026-03
Retires
2027-07-21
Open weights
no
Input
text, image, video, audio, pdf
Output
text
Price
US dollars per million tokens · google list price
Input
$0.30
Output
$2.50
Cache read
$0.03
Cache write
$0.0833
Batch and priority prices
US dollars per million tokens · per listing
| Price | Listing | $/M |
|---|---|---|
| Batch input | vertex_ai-language-models | $0.15 |
| Batch output | vertex_ai-language-models | $1.25 |
| Priority input | vertex_ai-language-models | $0.54 |
| Priority output | vertex_ai-language-models | $4.50 |
Batch input is 50% below the interactive input price at the same listing.
Rate limits
as listed
Requests per minute
15
Tokens per minute
250,000
Quality
7 benchmarks · 2 sources
Composite
39th
percentile of 334 scored models
Rank
204 / 334
±0.09 sd
Evidence
7 × 2
3 or more benchmarks, from 2 or more sources
Effort range
8.0
points between effort settings
reasoning69th3/4 bench
math32nd3/7 bench
general2nd1/1 bench
By reasoning effort
every setting placed on the same scale as the leaderboard
| Setting | Composite | Benchmarks behind it |
|---|---|---|
| low | 65th | 4 |
| minimal | 59th | 4 |
| high | 43rd | 6 |
Every score
one row per source and configuration — nothing averaged away
| Benchmark | Score | Configuration | Source | Run |
|---|---|---|---|---|
| GPQA diamond | 83.3 ±2.7 | high | Epoch AI | 2026-08-06 |
| ↳ | 75.8 ±3.1 | low | Epoch AI | 2026-08-06 |
| ↳ | 74.2 ±3.1 | minimal | Epoch AI | 2026-08-06 |
| Settings differ by 9.1 points on this benchmark. | ||||
| Chess Puzzles | 22.0 ±4.2 | high | Epoch AI | 2026-08-06 |
| ↳ | 21.0 ±4.1 | minimal | Epoch AI | 2026-08-06 |
| ↳ | 18.0 ±3.9 | low | Epoch AI | 2026-08-06 |
| Settings differ by 4.0 points on this benchmark. | ||||
| OTIS Mock AIME 2024-2025 | 71.1 ±6.8 | high | Epoch AI | 2026-08-06 |
| ↳ | 60.0 ±7.4 | low | Epoch AI | 2026-08-06 |
| ↳ | 51.1 ±7.5 | minimal | Epoch AI | 2026-08-06 |
| Settings differ by 20.0 points on this benchmark. | ||||
| Mystery Game Puzzles | 19.0 ±3.9 | low | Epoch AI | 2026-08-05 |
| ↳ | 12.0 ±3.3 | minimal | Epoch AI | 2026-08-05 |
| Settings differ by 7.0 points on this benchmark. | ||||
| FrontierMath-Tiers-1-3-v2 | 26.0 ±2.6 | high | Epoch AI | 2026-08-02 |
| FrontierMath-Tier-4-v2 | 0.0 ±0.0 | high | Epoch AI | 2026-08-02 |
| LiveBench | 63.8 | high · 23 runs | LiveBench | 2026-06-25 |
The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.
Available from 29 hosts
| Host | In $/M | Out $/M | Cache rd | Cache wr | Context | Output | Retires |
|---|---|---|---|---|---|---|---|
| googlevendor | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| kilo | $0.15 | $1.25 | $0.015 | $0.0417 | 1M | 66K | — |
| 302ai | $0.30 | $2.50 | — | — | 1M | 66K | — |
| abacus | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| crossmodel | $0.30 | $2.50 | $0.03 | $0.30 | 1M | 66K | — |
| edenai | $0.30 | $2.50 | $0.03 | $0.0833 | 1M | 66K | — |
| gemini | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| google-vertex | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| impossibl | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| llmgateway | $0.30 | $2.50 | $0.03 | $0.0833 | 1M | 66K | — |
| llmgateway-providers · google-ai-studio | $0.30 | $2.50 | $0.03 | $0.0833 | 1M | 66K | — |
| llmgateway-providers · google-vertex | $0.30 | $2.50 | $0.03 | $0.0833 | 1M | 66K | — |
| merge-gateway | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| nano-gpt | $0.30 | $2.50 | $0.03 | $0.0833 | 1M | 66K | — |
| neon | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| ofox | $0.30 | $2.50 | $0.03 | $0.083 | 1M | 66K | — |
| opencode | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| openrouter | $0.30 | $2.50 | $0.03 | $0.0833 | 1M | 66K | — |
| opper | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| orcarouter | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| perplexity | $0.30 | $2.50 | $0.03 | — | — | — | — |
| pioneer | $0.30 | $2.50 | $0.03 | $0.30 | 1M | 65K | — |
| requesty | $0.30 | $2.50 | $0.03 | — | 1M | 66K | — |
| vercel | $0.30 | $2.50 | $0.03 | — | 1M | 65K | — |
| vertex_ai-language-models | $0.30 | $2.50 | $0.03 | — | 1M | 66K | 2027-07-21 |
| cortecs | $0.33 | $2.749 | $0.033 | — | 1M | 66K | — |
| databricks | $0.375 | $3.125 | $0.0375 | $0.375 | 1M | 66K | — |
| venice | $0.375 | $3.125 | $0.0375 | — | 1M | 66K | — |
| sap-ai-core | — | — | — | — | 1M | 66K | — |
Price history
append-only observations · a listing writes a row only when its price moves
| Date | Host | Field | $/M |
|---|---|---|---|
| 2026-09-02 | openrouter | Cache read | $0.03 |
| 2026-09-06 | openrouter | Cache read | $0.03 |
| 2026-09-02 | openrouter | Input | $0.30 |
| 2026-09-06 | openrouter | Input | $0.30 |
| 2026-09-02 | openrouter | Output | $2.50 |
| 2026-09-06 | openrouter | Output | $2.50 |
More from Google
most-hosted first