Usage
Token consumption against your current plan allowance, broken down by model.
Plan allowance
Monthly plan · renews Oct 18, 2026
41%of plan allowance used
412,884,000 of 1,000,000,000 tokens · 28 days remaining in this periodUsage
Tokens over the last 7 days
Usage by model
Input and output token consumption across the last 90 days.
| Model | Requests | Input tokens | Output tokens | Mean latency | Success | Spend |
|---|---|---|---|---|---|---|
| Claude Sonnet 5claude-sonnet-5 | 3,795 | 1,153,680 | 507,619 | 869 ms | 99.5% | $7.97 |
| DeepSeek V4.1 Flashdeepseek-v4.1-flash | 1,777 | 1,254,562 | 552,007 | 571 ms | 97.6% | $3.73 |
| GLM 5.2glm-5.2 | 1,728 | 1,016,064 | 447,068 | 1,159 ms | 99.8% | $3.63 |
| GLM 5.3 Flashglm-5.3-flash | 1,037 | 599,386 | 263,730 | 946 ms | 97.6% | $2.18 |
| GPT 5.6 Lunagpt-5.6-luna | 720 | 220,320 | 96,941 | 696 ms | 97.3% | $1.51 |
| DeepSeek V4 Prodeepseek-v4-pro | 511 | 348,502 | 153,341 | 348 ms | 97.6% | $1.07 |
- Claude Sonnet 5claude-sonnet-5 · 3,795 requests · 1,661,299 tokens · 99.5% success
- DeepSeek V4.1 Flashdeepseek-v4.1-flash · 1,777 requests · 1,806,569 tokens · 97.6% success
- GLM 5.2glm-5.2 · 1,728 requests · 1,463,132 tokens · 99.8% success
- GLM 5.3 Flashglm-5.3-flash · 1,037 requests · 863,116 tokens · 97.6% success
- GPT 5.6 Lunagpt-5.6-luna · 720 requests · 317,261 tokens · 97.3% success
- DeepSeek V4 Prodeepseek-v4-pro · 511 requests · 501,843 tokens · 97.6% success
How usage is measured
What counts toward the allowance and what does not.
- Prompt tokens include the system message, history and tool definitions sent with the request.
- Completion tokens include reasoning tokens when a model emits them.
- Requests rejected before reaching a provider — bad key, unknown model — do not consume the allowance.
- When the allowance is exhausted, requests return 429 until the period renews.