Model Pricing
| Model | Input | Cached Input | Output |
|---|---|---|---|
subconscious/tim-qwen3.6-27b | $0.50 / 1M | $0.05 / 1M | $3.50 / 1M |
tim-qwen3.5-4b, tim-qwen3.5-9b, tim-qwen3.5-122b-a10b, tim-nemotron3-nano, tim-nemotron3-super, and tim-kimi2.6. Pricing will be announced when each model launches.
Input tokens include your prompt, system message, and conversation history.
Output tokens include everything the model generates, including reasoning tokens when thinking mode is enabled.
Rate Limits
Default limits per organization:| Limit | Default |
|---|---|
| Tokens per minute | 1,000,000 |
| Requests per minute | 100 |
| Tokens per day | 10,000,000 |
Billing
Your account is billed based on token usage. You can add credits from your dashboard, and usage is deducted automatically as you make requests. If you’d like uninterrupted service, auto-pay is available to top up your balance when it runs low.Cost Examples
| Use Case | Input Tokens | Output Tokens | Cost |
|---|---|---|---|
| Short Q&A (100 queries) | ~50K | ~100K | ~$0.38 |
| Document summarization (10 docs) | ~500K | ~50K | ~$0.43 |
| Batch analysis (1,000 items) | ~2M | ~1M | ~$4.50 |
Other Ways to Use Subconscious
Want to explore other ways to use our models and inference system in production?- Dedicated Endpoints: Reserved compute with guaranteed capacity and custom rate limits
- On-Prem Deployment: Run our inference stack in your own cloud or data center
- Local Devices: Deploy models directly on workstations, laptops, and edge hardware