Text-to-Text

Evaluate language models on reasoning, writing, coding, instruction following, factual accuracy, long-context understanding, and multi-turn coherence across 8 leading LLMs.

Lobster Test

Standard benchmarks: MMLU, HumanEval, GSM8K, ARC, TruthfulQA, MT-Bench, plus custom VCL prompts.

Write a short story about a lobster who enters a dirt bike race

Scoring Rubric

Metric1 (Low)5 (High)
ReasoningWrong answer, no logicCorrect, clear chain of thought
Instruction FollowingIgnores constraintsFollows every constraint exactly
Creative WritingGeneric / boringVivid, original, engaging
Code GenerationDoesn't compileCorrect, efficient, idiomatic
Factual AccuracyMultiple hallucinationsFully accurate, verifiable
Long-ContextLoses threadCoherent across full context

Models Tracked

Claude Opus 4

API

Anthropic flagship

Claude Sonnet 4

API

Anthropic high-value tier

GPT-4o

API

OpenAI multimodal workhorse

GPT-o3

API

OpenAI reasoning model

Gemini 2.5 Pro

API

Google, 1M+ context

DeepSeek V3 / R1

API

R1 = reasoning variant; open-weight

Llama 4 Maverick

Open Weight

Meta, via providers

Grok 3

API

xAI, reasoning capabilities

Metrics Evaluated

ReasoningCode generationInstruction followingHallucination rateLong-context

Results Coming Soon

Benchmark results will be published here after the first monthly run. Scores, rankings, trend charts, and the lobster gallery — all in one place.