// alternatives · October 2026
Mercury 3 alternatives
Cheaper models that come close, budget substitutes, and the cheapest step up — chosen on published LiveBench scores and live list price, never on our opinion. Every model here supports tool calling, as Mercury 3 does.
LiveBench hasn't published a run for Mercury 3. The strongest benchmarked model at a similar or lower price is Claude Haiku 5.5, scoring 69.8 overall.
Mercury 3: no published LiveBench run, $0.162 per 1M blended tokens, about $65.00/mo for a RAG assistant.
Benchmarked models at or below its price
Mercury 3 has no published LiveBench run, so these are the highest-scoring benchmarked models priced no more than 25% above it — a known quantity to test against.
| Model | Overall | vs Mercury 3 | Blended / 1M | RAG assistant | Context |
|---|---|---|---|---|---|
| Claude Haiku 5.5anthropic 23% more expensive. Head to head → | 69.8 | — | $0.200 | $74.00/mo | 1M |
| GPT-6 Lunaopenai 23% more expensive. Head to head → | 64.6 | — | $0.200 | $74.00/mo | 1.1M |
| GLM 5.3 Flashz-ai 68% cheaper. Head to head → | 62.1 | — | $0.052−68% | $34.46/mo | 1.0M |
Mercury 3 alternatives FAQ
What is the best alternative to Mercury 3?
Claude Haiku 5.5. 23% more expensive. Every suggestion supports tool calling, like Mercury 3.
How alternatives are chosen
- Scores — LiveBench release 2026-10-25. Only models with a published run are suggested on quality; a missing model means “not evaluated”, not “worse”.
- Prices — live list prices from OpenRouter, blended 3:1 input:output. Monthly figures price a RAG assistant with the published cached-input rate where there is one.
- Same job — suggestions must support every capability Mercury 3 has among tool calling and image input. Other listings of the same model (free or batch tiers) are variants, not alternatives, and are left out.
A few points on a public benchmark can hide large differences on your own tasks. Shortlist here, then decide with an eval — how to build one.