📊 AI Token Usage
Total Tokens
—
Est. Cost
—
AI Calls
—
Avg per Call
—
By Provider & Model
Loading…
By Content Type
Loading…
Provider Quality Guide — for content generation tasks (slides, podcasts, quizzes, flashcards)
| Provider / Model | Quality | Speed | Cost / chapter | Best for |
|---|---|---|---|---|
| Ollama llama3.2 | ${stars(5)} 5/10 | 🐢 Slow (local GPU) | Free | Experimenting, privacy |
| Groq llama-3.3-70b | ${stars(7)} 7/10 | 🚀 Very fast | ~$1–3 | Quick drafts, high volume |
| GPT-4o-mini | ${stars(8)} 8/10 | ⚡ Fast | ~$3–6 | Best value cloud option |
| Claude Haiku 4.5 | ${stars(8)} 8/10 | ⚡ Fast | ~$10–12 | Reliable quality, low cost |
| Claude Sonnet 4.6 ⭐ Recommended | ${stars(9)} 9/10 | ✅ Medium | ~$30–40 | Best balance quality/cost |
| GPT-4o | ${stars(9)} 9/10 | ✅ Medium | ~$40–60 | Top OpenAI quality |
| Claude Opus 4.8 | ${stars(10)} 10/10 | 🐌 Slow | ~$100–150 | Maximum quality, bulk not recommended |
Quality scores are for structured content generation (not general chat). Estimates per 270-page textbook chapter, all 4 content types.
Recent Calls
| Time | Provider | Model | Type | Input | Output | Total | Est. Cost |
|---|---|---|---|---|---|---|---|
| Loading… | |||||||