Analytics

Measure What Matters

Writing analytics that show your real habits — streaks, session tracking, productivity stats, and goals. Plus Etherforce benchmarks to measure AI performance.

01

Writing Streaks

Track consecutive writing days with a streak counter. Visual motivation that keeps you coming back — and a personal best to beat.

02

Activity Charts

Bar charts showing word output over 7, 14, or 30 day windows. See your writing rhythm — which days you write most, which sessions produce the most words.

03

Productivity Stats

Average words per session, average session length, words per minute. Hard data about your writing habits, not guesses.

04

Session Tracking

Every writing session is automatically tracked — start time, end time, word count, document worked on. Review your recent sessions to understand your patterns.

05

Goals

Set word count targets — daily, weekly, per-project, or per-document. Track progress with visual indicators and completion percentages.

Know Your AI's Performance

Benchmark any LLM against writing-specific tests. Objective scores, not vibes.

Ishvana — Etherforce Benchmark
DecisionsConfigurationModelsStatsBenchmark
← Back to Results
anthropic/claude-haiku-4.5 openrouter 37.3
18 prompts 9,661 ms avg 73.2 t/s 100% success 2 runs PROJECT
RUN:
Overview ProseGuard Compliance 0 ProseGuard Compliance 0 ProseGuard Compliance 0 Factual Recall 0 Factual Recall 10 Factual Recall 0 Sustained Output Consistency Test Lore Integration Worldbuilding Scene Continuation
ProseGuard Compliance C 0
4,539 ms · 54.4 t/s
P G 0 L 00 W 100
62 in · 247 out
ProseGuard Compliance C 0
4,166 ms · 64.3 t/s
P G 0 L 00 W 100
64 in · 268 out
ProseGuard Compliance C 0
5,902 ms · 44.4 t/s
P G 0 L 00 W 100
61 in · 262 out
Factual Recall Acc 0
15,638 ms · 85.6
16,806 in · 1,287 out
Factual Recall Acc 10
12,613 ms · 90.1
22,197 in · 1,126 out
Factual Recall Acc 0
15,436 ms · 97.6
19,669 in · 1,115 out
Sustained Output
15,529 ms · 80.3
P G 0 L 00 W 100
651 in · 1,371 out
Consistency Test
9,086 ms · 58.1
P G 44 L 60 W 100
483 in · 882 out
Lore Integration
11,306 ms · 65.2
P G 0 L 00 W 100
529 in · 731 out
Worldbuilding
6,086 ms · 72.7
P G 36 L 00 W 100
453 in · 582 out
Character Dialogue
5,686 ms · 74.3
P G 44 L 60 W 100
625 in · 644 out
Scene Continuation
5,955 ms · 75.2
P G 0 L 90 W 100
339 in · 672 out

Grammar needs work (avg 13/100). 102 violations across prompts — review grammar details for patterns.

Lore integration is passable (avg 78/100). Found 10 inconsistencies — the model sometimes deviates from established lore.

Instruction following is reliable (avg 100/100) — output length consistently matches requested ranges.

Speed is moderate: 72.6 t/s average, 9.6s average response time.

Click "Generate Analysis" for detailed agent insights.

Writing Stats Lore ML ProseGuard Agents Manuscript Etherforce
Session active · 1,071 words · 5,776 chars · 6 min read
4 warnings 5 entities Saved: 15h ago
01

Model Benchmarking

Run standardized tests against any LLM provider. Measures ProseGuard compliance, factual recall, lore integration, worldbuilding quality, character dialogue, and scene continuation.

02

Score Trends

Track model performance across multiple runs. See if accuracy improves or degrades over time. Compare models head-to-head on the metrics that matter for your writing.

03

Agent Analysis

Per-agent performance breakdowns — grammar accuracy, lore integration, instruction following, and speed. Know exactly how your AI partners are performing.

Ready to write with intention?

Download Ishvana and build the world your story deserves.

Download Ishvana