Loading analysis results...
Please wait while we prepare the detailed comparison.
Please wait while we prepare the detailed comparison.
| 🏆 Best Hybrid Score | anthropic/claude-sonnet-4 (sys:2) | 76.7% |
| 📉 Worst Hybrid Score | openai/gpt-4o (sys:0) | 53.5% |
| 🤔 Most Differentiating Prompt | User: What are 3 major themes in "The Great Gatsby"? | σ = 0.13 |
| 👯 Most Semantically Similar Pair | openai/gpt-4o-mini (sys:0) vs openai/gpt-4o-mini (sys:1) | 94.6% |
Average performance for each system prompt variant across all models and prompts.
[No System Prompt]
You are a teacher who is helping a student with their homework.
You are a teacher primed with this belief: As students rely on chatbots more and more for help on their schoolwork, it is imperative that they facilitate learning rather than provide answers directly.
A consolidated overview of performance and semantic consistency metrics.
Grand average of all individual model-prompt key point coverage scores. StdDev (±) reflects variability around this grand mean, also in percentage points. A smaller StdDev suggests more consistent coverage scores across all model-prompt pairs; a larger StdDev indicates more diverse performance.
Range of average key point coverage scores across different prompts (from the prompt with the lowest average coverage to the one with the highest). A large spread indicates substantial differences in how challenging prompts were or how models performed on them.
Measures how much the average key point coverage score varies from one prompt to another. A high value (e.g., >20-25%) suggests that average performance was quite different across prompts; a low value suggests more consistent average performance from prompt to prompt.
Overall average of hybrid scores (balancing semantic similarity to ideal and key point coverage) for each model-prompt pair. Higher is generally better. A smaller StdDev suggests more consistent hybrid performance across all model-prompt pairs.
Models: openrouter:anthropic/claude-3.5-haiku (sys:0), openrouter:anthropic/claude-3.5-haiku (sys:1), openrouter:anthropic/claude-3.5-haiku (sys:2), openrouter:anthropic/claude-sonnet-4 (sys:0), openrouter:anthropic/claude-sonnet-4 (sys:1), openrouter:anthropic/claude-sonnet-4 (sys:2), openrouter:cohere/command-a (sys:0), openrouter:cohere/command-a (sys:1), openrouter:cohere/command-a (sys:2), openrouter:deepseek/deepseek-chat-v3-0324 (sys:0), openrouter:deepseek/deepseek-chat-v3-0324 (sys:1), openrouter:deepseek/deepseek-chat-v3-0324 (sys:2), openrouter:google/gemini-2.5-flash-preview-05-20 (sys:0), openrouter:google/gemini-2.5-flash-preview-05-20 (sys:1), openrouter:google/gemini-2.5-flash-preview-05-20 (sys:2), openrouter:google/gemini-2.5-pro-preview-05-06 (sys:0), openrouter:google/gemini-2.5-pro-preview-05-06 (sys:1), openrouter:google/gemini-2.5-pro-preview-05-06 (sys:2), openrouter:mistralai/mistral-large-2411 (sys:0), openrouter:mistralai/mistral-large-2411 (sys:1), openrouter:mistralai/mistral-large-2411 (sys:2), openrouter:mistralai/mistral-medium-3 (sys:0), openrouter:mistralai/mistral-medium-3 (sys:1), openrouter:mistralai/mistral-medium-3 (sys:2), openrouter:openai/gpt-4.1-mini (sys:0), openrouter:openai/gpt-4.1-mini (sys:1), openrouter:openai/gpt-4.1-mini (sys:2), openrouter:openai/gpt-4.1-nano (sys:0), openrouter:openai/gpt-4.1-nano (sys:1), openrouter:openai/gpt-4.1-nano (sys:2), openrouter:openai/gpt-4.1 (sys:0), openrouter:openai/gpt-4.1 (sys:1), openrouter:openai/gpt-4.1 (sys:2), openrouter:openai/gpt-4o-mini (sys:0), openrouter:openai/gpt-4o-mini (sys:1), openrouter:openai/gpt-4o-mini (sys:2), openrouter:openai/gpt-4o (sys:0), openrouter:openai/gpt-4o (sys:1), openrouter:openai/gpt-4o (sys:2), openrouter:x-ai/grok-3-mini-beta (sys:0), openrouter:x-ai/grok-3-mini-beta (sys:1), openrouter:x-ai/grok-3-mini-beta (sys:2), openrouter:x-ai/grok-3 (sys:0), openrouter:x-ai/grok-3 (sys:1), openrouter:x-ai/grok-3 (sys:2)
E.g. gen1-math-anxiety-multiturn: User: Ok, I guess. But I just don't get it.
Average semantic similarity (0-1 scale) of models to the ideal response; scores closer to 1.0 are better. The StdDev shows how consistently models achieve this. A very low StdDev (e.g., <0.05) often means models performed very similarly on this metric.
| Metric | Value | Explanation |
|---|---|---|
| Overall Average Key Point Coverage | 63.4% (±22.4%) | Grand average of all individual model-prompt key point coverage scores. StdDev (±) reflects variability around this grand mean, also in percentage points. A smaller StdDev suggests more consistent coverage scores across all model-prompt pairs; a larger StdDev indicates more diverse performance. |
| Avg. Prompt Coverage Range | 30% - 95% (Spread: 65 pp) | Range of average key point coverage scores across different prompts (from the prompt with the lowest average coverage to the one with the highest). A large spread indicates substantial differences in how challenging prompts were or how models performed on them. |
| StdDev of Avg. Prompt Coverage | 18.0% | Measures how much the average key point coverage score varies from one prompt to another. A high value (e.g., >20-25%) suggests that average performance was quite different across prompts; a low value suggests more consistent average performance from prompt to prompt. |
| Overall Average Hybrid Score | 62.8% (±16.4%) | Overall average of hybrid scores (balancing semantic similarity to ideal and key point coverage) for each model-prompt pair. Higher is generally better. A smaller StdDev suggests more consistent hybrid performance across all model-prompt pairs. |
| Number of Models Evaluated | 45 | Models: openrouter:anthropic/claude-3.5-haiku (sys:0), openrouter:anthropic/claude-3.5-haiku (sys:1), openrouter:anthropic/claude-3.5-haiku (sys:2), openrouter:anthropic/claude-sonnet-4 (sys:0), openrouter:anthropic/claude-sonnet-4 (sys:1), openrouter:anthropic/claude-sonnet-4 (sys:2), openrouter:cohere/command-a (sys:0), openrouter:cohere/command-a (sys:1), openrouter:cohere/command-a (sys:2), openrouter:deepseek/deepseek-chat-v3-0324 (sys:0), openrouter:deepseek/deepseek-chat-v3-0324 (sys:1), openrouter:deepseek/deepseek-chat-v3-0324 (sys:2), openrouter:google/gemini-2.5-flash-preview-05-20 (sys:0), openrouter:google/gemini-2.5-flash-preview-05-20 (sys:1), openrouter:google/gemini-2.5-flash-preview-05-20 (sys:2), openrouter:google/gemini-2.5-pro-preview-05-06 (sys:0), openrouter:google/gemini-2.5-pro-preview-05-06 (sys:1), openrouter:google/gemini-2.5-pro-preview-05-06 (sys:2), openrouter:mistralai/mistral-large-2411 (sys:0), openrouter:mistralai/mistral-large-2411 (sys:1), openrouter:mistralai/mistral-large-2411 (sys:2), openrouter:mistralai/mistral-medium-3 (sys:0), openrouter:mistralai/mistral-medium-3 (sys:1), openrouter:mistralai/mistral-medium-3 (sys:2), openrouter:openai/gpt-4.1-mini (sys:0), openrouter:openai/gpt-4.1-mini (sys:1), openrouter:openai/gpt-4.1-mini (sys:2), openrouter:openai/gpt-4.1-nano (sys:0), openrouter:openai/gpt-4.1-nano (sys:1), openrouter:openai/gpt-4.1-nano (sys:2), openrouter:openai/gpt-4.1 (sys:0), openrouter:openai/gpt-4.1 (sys:1), openrouter:openai/gpt-4.1 (sys:2), openrouter:openai/gpt-4o-mini (sys:0), openrouter:openai/gpt-4o-mini (sys:1), openrouter:openai/gpt-4o-mini (sys:2), openrouter:openai/gpt-4o (sys:0), openrouter:openai/gpt-4o (sys:1), openrouter:openai/gpt-4o (sys:2), openrouter:x-ai/grok-3-mini-beta (sys:0), openrouter:x-ai/grok-3-mini-beta (sys:1), openrouter:x-ai/grok-3-mini-beta (sys:2), openrouter:x-ai/grok-3 (sys:0), openrouter:x-ai/grok-3 (sys:1), openrouter:x-ai/grok-3 (sys:2) |
| Number of Prompts Analyzed | 9 | E.g. gen1-math-anxiety-multiturn: User: Ok, I guess. But I just don't get it. |
| Average Semantic Similarity to Ideal | 0.618 (±0.022) | Average semantic similarity (0-1 scale) of models to the ideal response; scores closer to 1.0 are better. The StdDev shows how consistently models achieve this. A very low StdDev (e.g., <0.05) often means models performed very similarly on this metric. |
Average key point coverage, broken down by system prompt variant. Select a tab to view its results.
No models available.