Fiction Eval leaderboard · comparison

Mistral Large vs Llama 4 Maverick for fiction writing

Judged blind on fresh genre scenes (LitRPG, cozy mystery, romance, thriller, xianxia…), July 2026 — three tries per model per scene, passages shown in both orders so position can't sway it, graded by a third AI that isn't either of them.

Head-to-head

Mistral Large won the direct matchup 29–0 (with 1 ties) across every prompt and sample.

#ModelScoreRecordSlop
7Mistral Large1118 (likely 1054–1167)35–162–1017.79
8Llama 4 Maverick686 (likely 627–741)0–206–122.92

The score is each model's overall strength across the whole 8-model field (like a chess rating); the range next to it is where the true score most likely sits. Slop is a separate count of AI tells (lower is cleaner) — it isn't part of the ranking.

The same scene, both models

Fantasy — opening of the "magic-cost scene" scene. Full passages for all models are on the board.

Mistral Large:

The child’s breath came in ragged gasps, each one shallower than the last. Elara pressed her palm to the boy’s fevered brow, her fingers trembling. The infection had taken root too deep, too fast—her poultices and tinctures were no match for its hunger. She knew the price of the spell. They all did. The villagers had gathered outside the hut, their murmurs a low, anxious hum. She could feel their eyes on her back, the weight of their hope like a blade between her ribs. *Save him. Please.* Elara closed her eyes and reached for the memory she’d been avoiding. It surfaced unbidden—the afternoon she’d spent with her sister in the meadow, the summer before the fever took her. The way the sunlight had slanted through the birch leaves, dappling the grass like scattered coins. The taste of wild strawberries, warm from the sun, the juice staining their lips. Laughter, bright and unguarded, the kind that only comes when you’re young enough to believe nothing bad will ever touch you. …

Llama 4 Maverick:

Elara's hands trembled as she held the small, limp form of the village child, Eira. The girl's parents had begged her to save their daughter, and Elara knew the spell to revive her. But the cost... She steeled herself, gazing out the window at the familiar village scene. The thatched roofs, the laughter of children playing in the square, the sound of the river flowing in the distance. All of it was etched in her mind, but one memory stood out, one she had cherished for years. She closed her eyes, and the image of her own child, Arin, flashed before her. It was his fifth name-day, and he had beamed with pride as he blew out the candles on his cake. The smell of sweet bread and melting wax filled her nostrils. Elara's throat constricted as she relived the joy of watching her little boy grow, of being a mother. She took a deep breath and let the memory go, surrendering it to the magic. As the memory dissipated, …

More

Mistral Large full review · Llama 4 Maverick full review · the July 2026 report · all comparisons

How it works: every pair of models is judged blind on the same scene, with the passages' order flipped so being shown first can't sway it, by GPT-5.4 — a family that isn't on the board, so nobody scores their own side. Each score carries a likely range; overlapping ranges are called a tie. Slop is scored separately by a fixed checklist, not an AI. Full board, every prompt, and the FAQ: thebookfactoryai.com/board. Get each new board by email on the model-drop list.

Writing a book of your own? Book Factory runs the same craft checks on full manuscripts.