// head_to_head
Mercury 2.5 vs MiniMax M3.1 Flash
Published LiveBench scores across all seven categories, live list pricing, context windows, and the measured cost of a point of capability — for both models, side by side.
Mercury 2.5 is the cheaper of the two; neither can be ranked on quality here.
Mercury 2.5 does not have a published LiveBench run, so this comparison covers price, context and declared capabilities only. A missing score means "not evaluated", not "worse" — the right way to separate these two is an eval on your own workload.
inception
Mercury 2.5
- Blended / 1M
- $0.068
- Context
- 260K
- Released
- Sep 8, 2026
- Overall score
- Not evaluated
minimax
MiniMax M3.1 Flash
- Blended / 1M
- $0.131
- Context
- 1.0M
- Released
- Oct 19, 2026
- Overall score
- 66.9
Specs and pricing
| Metric | Mercury 2.5 | MiniMax M3.1 Flash |
|---|---|---|
| LiveBench overall Weighted mean of the seven LiveBench category scores, 0–100. Higher is better. | — | 66.9 |
| Cost per point Measured benchmark spend divided by overall score — dollars per point of capability. | — | $0.0015 |
| Blended price / 1M 3:1 input:output mix, the usual shape of production traffic. | $0.068win | $0.131 |
| Input price / 1M | $0.040win | $0.075 |
| Output price / 1M | $0.150win | $0.300 |
| Cached input / 1M Price of an input token served from the prompt cache, where the provider publishes one. | $0.0040win | $0.0075 |
| Context window | 260K | 1.0Mwin |
| Max output tokens | 66K | 524Kwin |
Benchmarks by category
An overall score averages away the thing you probably care about. These are the seven LiveBench categories scored separately — Mercury 2.5 on top, MiniMax M3.1 Flash below, both out of 100.
What each one costs to run
Per-token prices are hard to feel. These are monthly list costs for both models across five workload shapes, using each provider's published cached-input rate where there is one.
| Workload | Mercury 2.5 | MiniMax M3.1 Flash |
|---|---|---|
| Support chatbot 1.2K in / 400 out × 200K requests | $19.01/mo | $37.14/mo |
| RAG assistant 8K in / 600 out × 100K requests | $26.60/mo | $51.00/mo |
| Coding agent 40K in / 4K out × 20K requests | $23.84/mo | $46.20/mo |
| Document extraction 20K in / 1.5K out × 50K requests | $49.45/mo | $94.13/mo |
| Bulk classification 500 in / 20 out × 5M requests | $97.00/mo | $183.75/mo |
Which should you pick?
You need to fit large documents in one call
MiniMax M3.1 Flash
Wider context window — 1.0M against 260K.
You are cost-constrained
Mercury 2.5
Cheaper on blended list price at $0.068 per million tokens.
Mercury 2.5 vs MiniMax M3.1 Flash FAQ
Which is better, Mercury 2.5 or MiniMax M3.1 Flash?
Mercury 2.5 is the cheaper of the two; neither can be ranked on quality here. Mercury 2.5 does not have a published LiveBench run, so this comparison covers price, context and declared capabilities only. A missing score means "not evaluated", not "worse" — the right way to separate these two is an eval on your own workload.
Is Mercury 2.5 cheaper than MiniMax M3.1 Flash?
Mercury 2.5 is cheaper. On a 3:1 input:output blend, Mercury 2.5 lists at $0.068 per million tokens and MiniMax M3.1 Flash at $0.131 — Mercury 2.5 is 1.9× cheaper. Input and output are priced separately — Mercury 2.5 charges $0.040 in and $0.150 out, MiniMax M3.1 Flash charges $0.075 and $0.300 — so the model that looks cheaper flips depending on how output-heavy your workload is.
Does Mercury 2.5 or MiniMax M3.1 Flash have a bigger context window?
MiniMax M3.1 Flash has the larger context window: 260K for Mercury 2.5 against 1.0M for MiniMax M3.1 Flash. Note that a window you can fill is not a window you should fill — retrieval quality usually degrades well before the limit, and you pay for every token you put in it.
Do Mercury 2.5 and MiniMax M3.1 Flash support prompt caching?
Both publish a cached-input rate: $0.0040 per million for Mercury 2.5 and $0.0075 for MiniMax M3.1 Flash, against full input rates of $0.040 and $0.075. On a workload with a long stable prefix — a system prompt, a tool schema, a retrieved corpus — that changes the economics more than the headline price does.
Related comparisons
How these numbers are produced
- Price — provider list price from OpenRouter, refreshed every 15 minutes. “Blended” is a 3:1 input:output mix.
- Scores — LiveBench release 2026-10-25, using their own category map. Each model shows its strongest published run. A blank means “not evaluated”, never “bad”.
- Cost per point — the measured dollars LiveBench spent on the run, divided by the score it earned.
- “Win” — awarded only past a threshold: one full point on a benchmark score, 10% on a price, 25% on a context window. Anything tighter reports as a tie, because effort settings alone move a LiveBench score by more than that.
Published benchmarks rank models on someone else's tasks. Before committing, see LLM & agent evaluation for building an eval on your own.