Droid
Observed throughTokscale Coding Client Report
Factory Droid is an AI software-development agent for planning, implementing, and reviewing code.
Reality Snapshot
Observed tokens
21B-72.5%
over 30 days
Estimated tracked spend
$6.3K
Invoices may differ
Active developers
14-11
deduplicated over 30 days
Observed messages
1.5K
over 30 days
Most-used model
MiMo V2.5.061.2%
of observed tokens
Observed Usage
Observed tokensDaily active developers
Models used
Observed 30-day client usage by model
| Model | Tokens | Share | Change |
|---|---|---|---|
| MiMo V2.5.0Observed model | 12.8B | 61% | 0% |
| GLM 5.2-21Z Ai | 2.8B | 13% | 0% |
| OtherObserved model | 1.3B | 6.3% | 0% |
| GLM 5.2Z Ai | 1.2B | 6% | 0% |
| gpt-5-6-sol-0OpenAI | 809.5M | 3.9% | 0% |
| MiniMax M3Minimax | 663.8M | 3.2% | 0% |
| gpt-5-6-sol-fast-0OpenAI | 308.4M | 1.5% | 0% |
| claude-opus-5-2Anthropic | 277.4M | 1.3% | 0% |
Token composition
Share of observed client tokens
Input15.6B75%
Cache read5.2B25%
Output75.6M<1%
Reasoning28.7M<1%
Client Economics
MiMo V2.5.0$661%
GLM 5.2-21$1.8K28%
Other$1.3K21%
GLM 5.2$1.7K26%
gpt-5-6-sol-0$79913%
MiniMax M3$641%
gpt-5-6-sol-fast-0$3215.1%
claude-opus-5-2$0.000%
Estimated spend per 1M tracked tokens$0.30
Cache-read share25.1%
Spend basisEstimated from tracked model usage
Estimated tracked spend uses Tokscale model and provider pricing data. Actual invoices may differ.
Client Momentum
-72.5%30-day growthObserved coding client
Compared with the prior non-overlapping 30-day window
Provider footprint
- Observed model14.1B
- Z Ai4.1B
- OpenAI1.1B
- Minimax663.8M
- Anthropic277.4M
Client details
Canonical clientdroid
TelemetryAggregated Tokscale submissions
Current window2026-07-12 to 2026-08-10
Comparison window2026-06-12 to 2026-07-11
SpendEstimated; invoices may differ
About this client
Factory Droid is an AI software-development agent for planning, implementing, and reviewing code.
Links & sources
TOKSCALE INSIGHT
Droid recorded 21B coding tokens across 14 developers during the latest 30-day Tokscale window.
Evidence confidenceStable observed sample