LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

GPT-4 vs Mercury 2

Updated October 21, 2026. 23 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace GPT-4 Replace Mercury 2
GPT-4OpenAI
Coding
22.4
Overall
6.7
Output $/M
$60.00
0
Mercury 2Inception
Coding
0.0
Overall
10.8
Output $/M
$0.75
23
vs boards won of 23 shared

Decision reading

Pick GPT-4 for coding: it scores 22.4 on the Vibe Coding Index against 0.0 for Mercury 2.

Pick Mercury 2 when output tokens cost 80.0x less, or when the context window is larger, 128K against 8K tokens.

Model AGPT-4OpenAI22.4/1001 independent benchmarkNo cost per solved taskModel BMercury 2Inception0.0/1001 independent benchmarkNo cost per solved task
Shared results
23
GPT-4 only
22
Mercury 2 only
19
Like-for-like categories
5 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
GPT-4 22.4 from 1 benchmark, Mercury 2 0.0 from 1 benchmark.
Widest category
Reasoning, 32.8 points apart on average over 4 shared benchmarks.
Widest single gap
GPQA Diamond (Artificial Analysis run): GPT-4 34.9% against Mercury 2 77.0%, 42.1 points apart on a 0 to 100 scale.
Output price
GPT-4 $60.00 against Mercury 2 $0.75 per million output tokens, 80.0x apart.
Context window
GPT-4 takes 8K tokens, Mercury 2 128K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryGPT-4BarsMercury 2Ties
Coding03
Intelligence01
Reasoning04
Knowledge03
General012
All shared boards023
CodingLike-for-likeFour of these boards feed the Vibe Coding Index03
Coding benchmarks for GPT-4 and Mercury 2
BenchmarkGPT-4Mercury 2Leader
SWE-bench Verified22.4%Not measuredOnly GPT-4 measured
SWE-bench Lite39.7%Not measuredOnly GPT-4 measured
SWE-bench (full test set)12.5%Not measuredOnly GPT-4 measured
LMArena WebDevNot measured1171Only Mercury 2 measured
AA Coding Index13.131.1Mercury 2
SciCodeNot measured37.7%Only Mercury 2 measured
Terminal-Bench Hard (Artificial Analysis run)Not measured26.5%Only Mercury 2 measured
WeirdMLNot measured43.2%Only Mercury 2 measured
ALE-BenchNot measured785.6Only Mercury 2 measured
LMArena Text: coding12091390Mercury 2
LMArena Text: software and IT11951385Mercury 2
LMArena WebDev: ReactNot measured1160Only Mercury 2 measured
LMArena WebDev: HTMLNot measured1199Only Mercury 2 measured
Neither model has a result on the other 25 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for GPT-4 and Mercury 2
BenchmarkGPT-4Mercury 2Leader
AA Intelligence Index6.710.8Mercury 2
Epoch Capabilities Index125.7Not measuredOnly GPT-4 measured
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
Agentic work benchmarks for GPT-4 and Mercury 2
BenchmarkGPT-4Mercury 2Leader
AA Agentic IndexNot measured1.6Only Mercury 2 measured
τ²-Bench TelecomNot measured70.8%Only Mercury 2 measured
Neither model has a result on the other 37 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index04
Reasoning benchmarks for GPT-4 and Mercury 2
BenchmarkGPT-4Mercury 2Leader
GPQA Diamond (Artificial Analysis run)34.9%77.0%Mercury 2
Humanity's Last Exam (Artificial Analysis run)Not measured17.1%Only Mercury 2 measured
BIG-Bench Hard75.1%Not measuredOnly GPT-4 measured
LMArena Text: hard prompts12001362Mercury 2
LMArena Text: expert11491358Mercury 2
LMArena Text: hard prompts in English12091374Mercury 2
Neither model has a result on the other 15 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index3 boards, none shared
Math benchmarks for GPT-4 and Mercury 2
BenchmarkGPT-4Mercury 2Leader
GSM8K92.0%Not measuredOnly GPT-4 measured
LMArena Text: math1230Not measuredOnly GPT-4 measured
LMArena Text: mathematical work1228Not measuredOnly GPT-4 measured
Neither model has a result on the other 10 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index03
Knowledge benchmarks for GPT-4 and Mercury 2
BenchmarkGPT-4Mercury 2Leader
MMLU Pro (Artificial Analysis run)56.2%Not measuredOnly GPT-4 measured
MMLU86.4%Not measuredOnly GPT-4 measured
AI2 Reasoning Challenge96.3%Not measuredOnly GPT-4 measured
LMArena Text: science11771343Mercury 2
LMArena Text: medicine and healthcare11451363Mercury 2
LMArena Text: legal and government1191Not measuredOnly GPT-4 measured
LMArena Text: business and finance11451349Mercury 2
Neither model has a result on the other 10 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
Web design benchmarks for GPT-4 and Mercury 2
BenchmarkGPT-4Mercury 2Leader
Design Arena: websitesNot measured1036Only Mercury 2 measured
Design Arena: UI componentsNot measured998Only Mercury 2 measured
Design Arena: game devNot measured985Only Mercury 2 measured
Design Arena: data vizNot measured1009Only Mercury 2 measured
Design Arena: 3DNot measured982Only Mercury 2 measured
Design Arena: all code categoriesNot measured1022Only Mercury 2 measured
Design Arena: SVGNot measured998Only Mercury 2 measured
Design Arena: ASCII artNot measured990Only Mercury 2 measured
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index012
General benchmarks for GPT-4 and Mercury 2
BenchmarkGPT-4Mercury 2Leader
IFBench33.2%69.8%Mercury 2
Long context reasoning (AA-LCR)Not measured43.7%Only Mercury 2 measured
DTBench62.7%Not measuredOnly GPT-4 measured
LMCA17.1%Not measuredOnly GPT-4 measured
HellaSwag95.3%Not measuredOnly GPT-4 measured
WinoGrande87.5%Not measuredOnly GPT-4 measured
LMArena Text12061354Mercury 2
LMArena Text: instruction following12011328Mercury 2
LMArena Text: creative writing11911286Mercury 2
LMArena Text: English12201378Mercury 2
LMArena Text: non-English11941330Mercury 2
LMArena Text: Chinese11841414Mercury 2
LMArena Text: German1198Not measuredOnly GPT-4 measured
LMArena Text: Japanese1137Not measuredOnly GPT-4 measured
LMArena Text: Korean1087Not measuredOnly GPT-4 measured
LMArena Text: French1220Not measuredOnly GPT-4 measured
LMArena Text: Spanish1197Not measuredOnly GPT-4 measured
LMArena Text: Russian11951304Mercury 2
LMArena Text: Polish1265Not measuredOnly GPT-4 measured
LMArena Text: multi-turn12061357Mercury 2
LMArena Text: longer queries11901329Mercury 2
LMArena Text: writing and language12191300Mercury 2
LMArena Text: entertainment and media11841314Mercury 2
Neither model has a result on the other 23 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-4 and Mercury 2
BenchmarkGPT-4Mercury 2Leader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

GPT-4$0.060Fits in one request
Mercury 2$0.0006Fits in one request

Mercury 2 costs less

Repository review

50K input + 3K output tokens

GPT-4$1.68Exceeds the context window
Mercury 2$0.015Fits in one request

Mercury 2 costs less

Agent loop

200K input + 10K output tokens

GPT-4$6.60Exceeds the context window
Mercury 2$0.058Exceeds the context window

Mercury 2 costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of GPT-4 and Mercury 2
SpecificationOpenAI GPT-4Inception Mercury 2
MakerOpenAIInception
WeightsNo public weights listedNo public weights listed
ReleasedMay 28, 2023March 4, 2026
Context window8K128K
Input $/Mtok$30.00$0.25
Output $/Mtok$60.00$0.75
Coding index22.40.0
Independent measurements behind it11
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-4 better than Mercury 2 for coding?

Pick GPT-4 for coding: it scores 22.4 on the Vibe Coding Index against 0.0 for Mercury 2. Pick Mercury 2 when output tokens cost 80.0x less, or when the context window is larger, 128K against 8K tokens.

Which is cheaper, GPT-4 or Mercury 2?

GPT-4 costs $30.00 per million input tokens and $60.00 per million output tokens; Mercury 2 costs $0.25 and $0.75. Cost per solved task is known for neither of them.

Which has the bigger context window, GPT-4 or Mercury 2?

Mercury 2, with 128K tokens against 8K.

How many benchmarks is the GPT-4 vs Mercury 2 comparison based on?

23 benchmarks have a result for both models as of October 21, 2026. GPT-4's coding index rests on 1 independent benchmark and Mercury 2's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next