Google — pricing & performance
Models served
59
of 1,111 tracked models
Mean output speed
83
tokens per second, averaged over
45 measured endpoints
45 measured endpoints
Mean latency
4.5s
time to first token, averaged over
45 measured endpoints
45 measured endpoints
Cheapest blended
$0.087
per 1M tokens at a 3:1
input:output blend
input:output blend
Output speed on Google
Output tokens per second for the 45 of 59 models on Google with published throughput telemetry, the most recent figure OpenRouter publishes for each host. The full table below lists all 59, measured or not. Higher is better. We do not run these measurements ourselves.
The Known Good
59 models on Google
Full leaderboard →
| Model | Creator | Input $/M | Output $/M | Blended | Speed | TTFT | Context |
|---|---|---|---|---|---|---|---|
|
|
Anthropic | $10.00 | $50.00 | $20.00 | 48 | 4.5s | 1.00M |
|
|
Anthropic | $10.00 | $50.00 | $20.00 | 47 | 4.8s | 1.00M |
|
|
Anthropic | $1.00 | $5.00 | $2.00 | 72 | 0.9s | 200k |
|
|
Anthropic | $15.00 | $75.00 | $30.00 | 24 | 2.6s | 200k |
|
|
Anthropic | $15.00 | $75.00 | $30.00 | 3 | 1.8s | 200k |
|
|
Anthropic | $5.00 | $25.00 | $10.00 | 43 | 1.0s | 200k |
|
|
Anthropic | $5.00 | $25.00 | $10.00 | 43 | 2.0s | 1.00M |
|
|
Anthropic | $5.00 | $25.00 | $10.00 | 55 | 1.5s | 1.00M |
|
|
Anthropic | $5.00 | $25.00 | $10.00 | 67 | 4.8s | 1.00M |
|
|
Anthropic | $3.00 | $15.00 | $6.00 | 41 | 1.4s | 1.00M |
|
|
Anthropic | $3.00 | $15.00 | $6.00 | 35 | 1.6s | 1.00M |
|
|
Anthropic | $3.00 | $15.00 | $6.00 | 42 | 2.0s | 1.00M |
|
|
Anthropic | $2.00 | $10.00 | $4.00 | 71 | 3.7s | 1.00M |
|
|
Anthropic | $5.00 | $25.00 | $10.00 | 59 | 3.6s | 1.00M |
|
|
DeepSeek | $0.60 | $1.70 | $0.88 | — | — | 164k |
|
|
DeepSeek | $0.56 | $1.68 | $0.84 | 25 | 0.9s | 164k |
|
|
$0.30 | $2.50 | $0.85 | 97 | 3.1s | 1.05M | |
|
|
$0.15 | $1.25 | $0.42 | — | — | 1.05M | |
|
|
$0.10 | $0.40 | $0.17 | 72 | 0.6s | 1.05M | |
|
|
$0.050 | $0.20 | $0.087 | — | — | 1.05M | |
|
|
$1.25 | $10.00 | $3.44 | 112 | 17.3s | 1.05M | |
|
|
$0.62 | $5.00 | $1.72 | — | — | 1.05M | |
|
|
$1.25 | $10.00 | $3.44 | 112 | 17.3s | 1.05M | |
|
|
$1.25 | $10.00 | $3.44 | 112 | 17.3s | 1.05M | |
|
|
$0.25 | $1.50 | $0.56 | 86 | 71.0s | 1.05M | |
|
|
$0.25 | $1.50 | $0.56 | — | — | 1.05M | |
|
|
$0.12 | $0.75 | $0.28 | 68 | 2.8s | 1.05M | |
|
|
$0.12 | $0.75 | $0.28 | — | — | 1.05M | |
|
|
$1.00 | $6.00 | $2.25 | 98 | 2.9s | 1.05M | |
|
|
$1.00 | $6.00 | $2.25 | — | — | 1.05M | |
|
|
$0.75 | $4.50 | $1.69 | 155 | 1.4s | 1.05M | |
|
|
$0.75 | $4.50 | $1.69 | — | — | 1.05M | |
|
|
$0.15 | $1.25 | $0.42 | 132 | 1.7s | 1.05M | |
|
|
$0.15 | $1.25 | $0.42 | — | — | 1.05M | |
|
|
$0.38 | $1.88 | $0.75 | 116 | 1.5s | 1.05M | |
|
|
$0.38 | $1.88 | $0.75 | — | — | 1.05M | |
|
|
$0.38 | $1.88 | $0.75 | 82 | 2.2s | 1.05M | |
|
|
$0.38 | $1.88 | $0.75 | — | — | 1.05M | |
|
|
$0.38 | $1.88 | $0.75 | 87 | 1.5s | 1.05M | |
|
|
$0.38 | $1.88 | $0.75 | — | — | 1.05M | |
|
|
$0.15 | $0.60 | $0.26 | 28 | 1.4s | 262k | |
|
|
$0.30 | $2.50 | $0.85 | 111 | 1.1s | 33k | |
|
|
$0.25 | $1.50 | $0.56 | — | — | 131k | |
|
|
$0.50 | $3.00 | $1.12 | 114 | 8.9s | 131k | |
|
|
$0.25 | $1.50 | $0.56 | 232 | 3.6s | 66k | |
|
|
$1.00 | $6.00 | $2.25 | — | — | 66k | |
|
|
$2.00 | $12.00 | $4.50 | 60 | 2.7s | 66k | |
|
|
Meta | $0.72 | $0.72 | $0.72 | 71 | 0.3s | 128k |
|
|
Meta | $0.35 | $1.15 | $0.55 | 78 | 0.2s | 524k |
|
|
Meta | $0.25 | $0.70 | $0.36 | 59 | 0.5s | 1.31M |
|
|
MiniMax | $0.30 | $1.20 | $0.53 | 102 | 0.3s | 197k |
|
|
Moonshot AI | $0.60 | $2.50 | $1.07 | 70 | 0.8s | 262k |
|
|
OpenAI | $0.090 | $0.36 | $0.16 | 199 | 0.3s | 131k |
|
|
OpenAI | $0.070 | $0.25 | $0.12 | 267 | 0.3s | 131k |
|
|
Alibaba | $0.22 | $0.88 | $0.39 | 45 | 0.6s | 262k |
|
|
Alibaba | $0.22 | $1.80 | $0.61 | 46 | 1.0s | 262k |
|
|
Alibaba | $0.15 | $1.20 | $0.41 | 98 | 0.6s | 262k |
|
|
Alibaba | $0.15 | $1.20 | $0.41 | 45 | 0.3s | 262k |
|
|
Z.ai | $0.60 | $2.20 | $1.00 | 85 | 0.5s | 200k |
The Known Good
Source. Prices, endpoint availability and throughput for Google are ingested from OpenRouter;
model metadata is enriched from models.dev. A blank cell means the figure was not published for that endpoint —
it never means zero. Blended price is (3 × input + output) ÷ 4, the same 3:1 blend used everywhere on this
site; see methodology. Quality scores are not provider-specific:
a model scores the same wherever it is hosted, so quality lives on the
model pages.