Gemma 4 26B A4B
Google · Gemma · measured on live video
Overshoot API id google/gemma-4-26B-A4B-it
Real-time under 200 msOpen weightsImage + Video
- Rank
- #12of 13
- Intelligence
- 69
- Speed
- 429 tok/s
- TTFT
- 30 ms
- E2E
- 86 ms
- Blended $/1M
- $0.12
- Params
- 26B
- Context
- 128k
- License
- Gemma
- Released
- Apr 2026
Speed, TTFT, E2E, and price at Overshoot, US East (Virginia) reference region. Rank is position on the VLM Intelligence Index.
Capability breakdown
Intelligence over time
Provider & region options
Fastest loop: Overshoot · US East (Virginia) · 86 ms end to end.
| Overshoot | US East (Virginia) | 429 tok/s | 30 ms | 86 ms | $0.12 | <200 ms |
| Overshoot | US West (Oregon) | 429 tok/s | 43 ms | 99 ms | $0.12 | <200 ms |
| Overshoot | EU West (Ireland) | 429 tok/s | 50 ms | 105 ms | $0.12 | <200 ms |
| Overshoot | EU Central (Frankfurt) | 429 tok/s | 60 ms | 116 ms | $0.12 | <200 ms |
| Overshoot | Asia Pacific (Tokyo) | 429 tok/s | 67 ms | 123 ms | $0.12 | <200 ms |
| Overshoot | Asia Pacific (Mumbai) | 429 tok/s | 88 ms | 143 ms | $0.12 | <200 ms |
| Overshoot | Middle East (Dubai) | 429 tok/s | 94 ms | 150 ms | $0.12 | <200 ms |
| Together | US East (Virginia) | 261 tok/s | 58 ms | 150 ms | $0.12 | <200 ms |
| Overshoot | South America (São Paulo) | 429 tok/s | 101 ms | 157 ms | $0.12 | <200 ms |
| Together | US West (Oregon) | 261 tok/s | 65 ms | 157 ms | $0.12 | <200 ms |
| vLLM (self-host) | US East (Virginia) | 240 tok/s | 66 ms | 166 ms | $0.08 | <200 ms |
| Novita | US East (Virginia) | 241 tok/s | 69 ms | 168 ms | $0.09 | <200 ms |
| Together | EU West (Ireland) | 261 tok/s | 80 ms | 171 ms | $0.12 | <200 ms |
| vLLM (self-host) | EU West (Ireland) | 240 tok/s | 93 ms | 193 ms | $0.08 | <200 ms |
| Novita | Asia Pacific (Tokyo) | 241 tok/s | 99 ms | 198 ms | $0.09 | <200 ms |
Sample outputs
Illustrative samples shared across models, not verbatim output from Gemma 4 26B A4B.
“Describe what's happening in this camera frame.”
A forklift is reversing toward loading bay 3 while a worker in a hi-vis vest signals from the left. Two pallets remain unwrapped near the dock edge.
“Read the shipping label and return structured JSON.”
{ "tracking": "1Z A24 9W7 03 8421 990 4", "carrier": "UPS", "weight_kg": 12.4, "dest": "Portland, OR 97204" }
“Point to every person not wearing a hard hat.”
2 detections, [x:0.41, y:0.62, w:0.08, h:0.19] conf 0.94 · [x:0.77, y:0.55, w:0.07, h:0.21] conf 0.88.
Related models
Where it ranks
- #8Gemini 3 Flash75.3
- #9Claude Haiku 4.574.2
- #10Holo3 35B A3B74.0
- #11Gemma 4 31B73.0
- #12Gemma 4 26B A4B69.3
- #13GPT-5.4 nano63.6