Mercury Edit 2 API Pricing
Mercury Edit 2 carries the same card as Mercury 2 - $0.25/M input, $0.025/M cached, $0.75/M output - but it is not a chat model. It serves v1/fim/completions and v1/edit/completions: fill-in-the-middle autocomplete and next-edit suggestions for code tooling, with a 32K window and 8,192 max output tokens. Inception notes it stays supported for existing customers.
Run the numbers.
The rate is identical to Mercury 2, so the calculator differs only in the shape of the workload: editing traffic is many small requests, not a few large ones, and the per-request figures below matter more than the per-million rate.
Real-world presets.
Inline completion
Next-edit suggestion
Single-file edit
Multi-file refactor
Paste text. See tokens. See cost.
This is a chars-per-token approximation, not a real tokenizer. Actual tokens vary by language, code density, and tool-call overhead — counts are typically ±10–20% off for English prose, more for code or non-Latin scripts. For exact billing, use the vendor's official tokenizer.
| Model | Input /M | Output /M | Effective blended | Context | Best for |
|---|---|---|---|---|---|
| Mercury Edit 2 Current | $0.25 cache $0.025 | $0.75 | $0.12 agentic 92/8 | 32K | IDE autocomplete and next-edit suggestions |
| Mercury 2 | $0.25 cache $0.025 | $0.75 | $0.12 identical card, chat endpoint | 128K | General production traffic |
| Mercury 2.5 Preview | $0.20 cache $0.02 | $0.75 | $0.108 cheaper input, longer window | 260K | Reasoning at the lowest Mercury rate |
| Doubao Seed 2.0 Code | $0.476 cache $0.0952 | $2.38 | $0.341 cross-vendor coding row | 256K | IDE agents on Volcano Ark |
| KAT-Coder-Air V2.5 | $0.149 cache $0.0298 | $0.595 | $0.0947 cross-vendor coding row | 256K | Budget coding model |
Frequently asked.
Why it costs the same as the chat model, what the endpoints do, and what "supported for existing customers" means for a new integration.
Q · 01 What does Mercury Edit 2 cost? +
$0.25/M input, $0.025/M cached input, $0.75/M output, giving $0.12/M effective under our 92/8 blend. Inception does not discount the editing endpoints.Q · 02 How is it different from Mercury 2? +
v1/fim/completions and v1/edit/completions instead of chat completions, which is what code tooling needs for inline autocomplete and next-edit suggestions. The context window is 32K on both endpoints against Mercury 2's 128K, and max output is 8,192 against 50,000.