Scores over time

Best recorded score per model; higher is better. The line follows the running best. Evaluation contexts can differ.

F1 score scores over timeOne point per model's best recorded score, dated by reported date or model release date. A line shows the running best. The data table below contains every plotted point.9574.62026-04-082026-09-02Muse Spark: 74.8% · Reported 2026-04-08Claude Opus 5: 95.0% · Reported 2026-07-24Muse Glimmer 30B: 74.6% · Reported 2026-08-10Muse Spark 1.3: 90.3% · Reported 2026-09-02Muse Spark 1.2: 85.9% · Reported 2026-09-02
Data table
Scores plotted over time
ModelDateScore
Muse Spark (Reported)74.8%
Claude Opus 5 (Reported)95.0%
Muse Glimmer 30B (Reported)74.6%
Muse Spark 1.3 (Reported)90.3%
Muse Spark 1.2 (Reported)85.9%

Results

9 results

DeepSearchQA 900-prompt set results
ProviderModelScoreSourceRegistry No.
Meta AI (originally Facebook AI Research)Muse Spark 1.2 (xhigh)85.9%dev.meta.ai for Muse Spark 1.2 (xhigh) on DeepSearchQA 900-prompt set (opens in a new tab)F1 scoreReported Model pageMetaSelf-reportedRecord for Muse Spark 1.2 on DeepSearchQA 900-prompt set80003
Meta AI (originally Facebook AI Research)Muse Spark 1.3 (max)90.3%dev.meta.ai for Muse Spark 1.3 (max) on DeepSearchQA 900-prompt set (opens in a new tab)F1 scoreReported Model pageMetaSelf-reportedRecord for Muse Spark 1.3 on DeepSearchQA 900-prompt set80005
Meta AI (originally Facebook AI Research)Muse Glimmer 30B (high)74.6%dev.meta.ai for Muse Glimmer 30B (high) on DeepSearchQA 900-prompt set (opens in a new tab)F1 scoreReported Model pageMetaSelf-reportedRecord for Muse Glimmer 30B on DeepSearchQA 900-prompt set80004
AnthropicClaude Opus 5 (medium)92.8%www-cdn.anthropic.com for Claude Opus 5 (medium) on DeepSearchQA 900-prompt set (opens in a new tab)F1 scoreReported System cardAnthropicSelf-reportedRecord for Claude Opus 5 on DeepSearchQA 900-prompt set20014
AnthropicClaude Opus 5 (max)95.0%www-cdn.anthropic.com for Claude Opus 5 (max) on DeepSearchQA 900-prompt set (opens in a new tab)F1 scoreReported System cardAnthropicSelf-reportedRecord for Claude Opus 5 on DeepSearchQA 900-prompt set20014
AnthropicClaude Opus 5 (high)94.1%www-cdn.anthropic.com for Claude Opus 5 (high) on DeepSearchQA 900-prompt set (opens in a new tab)F1 scoreReported System cardAnthropicSelf-reportedRecord for Claude Opus 5 on DeepSearchQA 900-prompt set20014
AnthropicClaude Opus 5 (low)91.0%www-cdn.anthropic.com for Claude Opus 5 (low) on DeepSearchQA 900-prompt set (opens in a new tab)F1 scoreReported System cardAnthropicSelf-reportedRecord for Claude Opus 5 on DeepSearchQA 900-prompt set20014
AnthropicClaude Opus 5 (xhigh)94.6%www-cdn.anthropic.com for Claude Opus 5 (xhigh) on DeepSearchQA 900-prompt set (opens in a new tab)F1 scoreReported System cardAnthropicSelf-reportedRecord for Claude Opus 5 on DeepSearchQA 900-prompt set20014
Meta AI (originally Facebook AI Research)Muse Spark (thinking)74.8%ai.meta.com for Muse Spark (thinking) on DeepSearchQA 900-prompt set (opens in a new tab)F1 scoreReported Evaluation methodologyMetaSelf-reportedRecord for Muse Spark on DeepSearchQA 900-prompt set80001