Gemini 3.7 Flash Iterates the Cheap Workhorse — Pro Still Missing
Google released Gemini 3.7 Flash — 23 days after 3.6 Flash — claiming large coding and agent gains and an introductory $0.75/$3.75 per million tokens through year-end. Artificial Analysis scores 3.7 Flash (high) at 56 on the Intelligence Index, up 4 points from 3.6 Flash. Gemini 3.5 Pro still not shipped.
Google is iterating the cheap workhorse every three weeks while the Pro flagship slips — buying share with price and agent scores rather than a new top-end name. At $0.75/$3.75, 3.7 Flash undercuts many Sol/Fable API rungs for coding agents.
Google introduced Gemini 3.7 Flash, calling it the most intelligent Flash “workhorse” yet for coding and agents. The model arrives three weeks after Gemini 3.6 Flash (Jul 21) and is attributed to algorithmic innovations and developer feedback, not a larger base. Introductory API price: $0.75 / $3.75 per million input/output tokens through December 31, 2026, then $1.50 / $7.50. Gemini Spark (AI Pro/Ultra, 160+ countries, not EEA/UK/Nigeria/Switzerland) switches to 3.7 Flash the same day. Independent Artificial Analysis scores 3.7 Flash (high) at 56 on the Intelligence Index (+4 vs 3.6 Flash).
What Google published
| Item | Detail |
|---|---|
| ID | gemini-3.7-flash (stable, no preview suffix) |
| Author | Tulsee Doshi, Senior Director, Product Management |
| Vendor coding | FrontierCode 1.1 Main 43.6% vs 34.4%; DeepSWE v1.1 65.3% vs 49.0% (vs 3.6 Flash) |
| WebDev Arena (Arena.ai) | Elo 1588 vs 1538 |
| GDP.pdf | 34.0% vs 22.0% |
| AutomationBench | 30.4% vs 17.0% |
| Intro price | $0.75 / $3.75 per 1M in/out through Dec 31, 2026 |
| Safety | Updated CBRN + cyber-offense safeguards; model card at DeepMind |
Google did not ship the delayed Gemini 3.5 Pro flagship with this drop. Reuters and Ars noted the Flash cadence while Pro remains outstanding.
Independent rankings
| Source | Snapshot | Standing |
|---|---|---|
| Artificial Analysis Intelligence Index | AA article Aug 13, 2026 | 3.7 Flash (high) = 56 (+4 vs 3.6); just behind GPT-5.6 Terra (max, 57) and Muse Spark 1.2 (xhigh, 57). Medium 53; low 51 |
| Arena.ai WebDev Arena | Google-cited | 1588 Elo vs 3.6 Flash 1538 |
| LMArena overall text | Research window | Not treated as independently confirmed here beyond the WebDev citation |
Capability claim holds up independently (+4 AA). Pricing needs the asterisk some secondaries printed: Google also moved 3.6 Flash to the same intro rate the same day, so the two workhorses bill the same until year-end.
Limits
- Gemini 3.5 Pro still not shipped.
- 3.6 Flash moved to the same intro price — half-price is not unique to 3.7.
- LMArena overall text not independently confirmed beyond WebDev.
Sources
- Google Blog: Introducing Gemini 3.7 Flash (August 13, 2026)
- DeepMind model card: Gemini 3.7 Flash
- Artificial Analysis: Gemini 3.7 Flash on the Intelligence vs. Time per Task frontier (August 13, 2026)
- Reuters: Google unveils Gemini 3.7 Flash (August 13, 2026)
- Ars Technica: Gemini 3.7 Flash three weeks after previous release (August 13, 2026)