// head_to_head
Mercury 3 vs GPT-6.1 Terra
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 3 is the cheaper of the two; neither can be ranked on quality here.
Mercury 3 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 3
- Blended / 1M
- $0.162
- Context
- 1M
- Released
- Oct 16, 2026
- Overall score
- Not evaluated
openai
GPT-6.1 Terra
- Blended / 1M
- $2.00
- Context
- 1.1M
- Released
- Oct 26, 2026
- Overall score
- 77.8
Specs and pricing
| Metric | Mercury 3 | GPT-6.1 Terra |
|---|---|---|
| LiveBench overall Weighted mean of the seven LiveBench category scores, 0–100. Higher is better. | — | 77.8 |
| Cost per point Measured benchmark spend divided by overall score — dollars per point of capability. | — | $0.0067 |
| Blended price / 1M 3:1 input:output mix, the usual shape of production traffic. | $0.162win | $2.00 |
| Input price / 1M | $0.100win | $1.00 |
| Output price / 1M | $0.350win | $5.00 |
| Cached input / 1M Price of an input token served from the prompt cache, where the provider publishes one. | $0.010win | $0.050 |
| Context window | 1M | 1.1M |
| Max output tokens | 131K | 128K |
Benchmarks by category
An overall score averages away the thing you probably care about. These are the seven LiveBench categories scored separately — Mercury 3 on top, GPT-6.1 Terra 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 3 | GPT-6.1 Terra |
|---|---|---|
| Support chatbot 1.2K in / 400 out × 200K requests | $45.52/mo | $571.60/mo |
| RAG assistant 8K in / 600 out × 100K requests | $65.00/mo | $720.00/mo |
| Coding agent 40K in / 4K out × 20K requests | $57.60/mo | $668.00/mo |
| Document extraction 20K in / 1.5K out × 50K requests | $121.75/mo | $1,328/mo |
| Bulk classification 500 in / 20 out × 5M requests | $240.00/mo | $2,525/mo |
Which should you pick?
You are cost-constrained
Mercury 3
Cheaper on blended list price at $0.162 per million tokens.
Mercury 3 vs GPT-6.1 Terra FAQ
Which is better, Mercury 3 or GPT-6.1 Terra?
Mercury 3 is the cheaper of the two; neither can be ranked on quality here. Mercury 3 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 3 cheaper than GPT-6.1 Terra?
Mercury 3 is cheaper. On a 3:1 input:output blend, Mercury 3 lists at $0.162 per million tokens and GPT-6.1 Terra at $2.00 — Mercury 3 is 12× cheaper. Input and output are priced separately — Mercury 3 charges $0.100 in and $0.350 out, GPT-6.1 Terra charges $1.00 and $5.00 — so the model that looks cheaper flips depending on how output-heavy your workload is.
Does Mercury 3 or GPT-6.1 Terra have a bigger context window?
They are effectively the same — 1M for Mercury 3 and 1.1M for GPT-6.1 Terra.
Do Mercury 3 and GPT-6.1 Terra support prompt caching?
Both publish a cached-input rate: $0.010 per million for Mercury 3 and $0.050 for GPT-6.1 Terra, against full input rates of $0.100 and $1.00. 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.