API v0.1 · Live

Emotions aren't simple.
Now you can measure their complexity.

Give your AI the ability to tell guilt from love, and respond differently to each.

Fossier scores the complexity of any emotion pair on a 0–198 scale. 28 named states, from guilt to ambivalence, each with an empathic difficulty level. Built on 6 years of research, validated on 28,218 texts.

---Pearson r (XED)
---Texts validated
---Named dyads
< ---p-value
fossier-api
0
Pleasure · Arousal · Dominance
0
Not a weekend project
0
Learned, not hardcoded
0
Green before every ship
Early Access · Beta Built on 6 years of research. Validated on 28,218 texts from 2 independent datasets. Early adopters get locked-in pricing →
Capabilities

Built on six years of research

Weighted PAD Distance

Emotional distance computed in 3D Pleasure-Arousal-Dominance space with cross-validated axis weights. Not a simple sentiment score.

V 1.30
A 0.74
D 0.96

28 Named Dyads

Every emotional pair mapped and named, from Mepris to Culpabilite.

MeprisOptimisme CulpabiliteDesespoir AmourAnxiete +22 more

3 Empathic Levels

Complexity maps to cognitive demand for empathic response.

Instinctive
Contextuelle
Creation

Validated at Scale

Pearson correlation between predicted complexity and observed co-occurrence across two independent multilabel datasets.

-0.854 XED
Why Fossier

This is not sentiment analysis

Sentiment Analysis

  • Binary: positive or negative
  • Joy + Fear = "mixed" (unhelpful)
  • No distance between emotions
  • Flat output, no empathic context
vs

Fossier API

  • 0–198 complexity scale
  • Joy + Fear = Culpabilite (score: 148)
  • Weighted 3D PAD distance
  • 3 empathic levels + 28 named dyads
Use Cases

What changes when AI understands complexity

The NPC Who Mourns

A player approaches an NPC whose wife just died. They say: "I'm sorry for your loss."
Without Fossier The NPC responds: "Thank you." Generic. Flat. Forgettable.
With Fossier The NPC clenches his fists: "She didn't deserve this." Joy+Tristesse = Ambivalence (C=143, creation). The AI knows this moment demands a response that holds both grief and love at once.

The Chatbot That Knows When To Stop

A customer writes: "I trusted you and you closed my account without warning."
Without Fossier Bot responds with template apology: "We're sorry for the inconvenience." Completely tone-deaf.
With Fossier Confiance+Degout = Ambivalence (C=137, contextuelle). Score triggers human escalation. A real person handles the trust rupture, not a template.

The Script That Finds Depth

A streaming platform analyzes 10,000 pilot scripts to find which ones have real emotional depth.
Without Fossier Sentiment analysis flags everything as "mixed" or "emotional." No ranking, no signal.
With Fossier Script A: avg C = 142 (creation). Script B: avg C = 67 (instinctive). The number tells you which script makes you feel something.
Live Scenarios

Feel the score. Click a story.

Joie Tristesse

The Bittersweet Goodbye

Your best friend is moving abroad. You're happy for them. You're devastated for yourself. Both feelings are real, and neither cancels the other.

0

Maximum score possible. This is why "bittersweet" exists as a word. Some feelings won't fit in a single label.

Confiance Degout

The Betrayal

You discover your business partner has been siphoning funds for months. You trusted them completely. Now every shared memory feels contaminated.

0

Trust + Disgust is cognitively disorienting. Your brain can't reconcile "I believed in you" with "you make me sick."

Joie Peur

The Soldier's Return

He survived and came home. She's overjoyed. But every car backfire makes him flinch, and she's terrified it will never be the same.

0

Surprise: Joy + Fear = Guilt in Plutchik's model. Survivor's guilt isn't sadness. It's relief and dread crashing into each other.

Joie Colere

Graduation Day

She walks across the stage with honors. Her parents didn't come. She proved everyone wrong. And she's furious she had to.

0

Joy + Anger = Pride in Plutchik's taxonomy. An unexpected tertiary dyad. Pride is never simple.

Degout Colere

The Easy One

Someone cuts in line and spits on the floor. You feel exactly what you'd expect. No ambiguity, no inner conflict.

0

Low score = low cognitive load. Your brain processes this instantly. Compare with 188 above. That gap is what Fossier measures.

Start Free · 100 requests/day

No credit card required. Full model access.

Playground

Compute emotional complexity in real time

1. First emotion

2. Second emotion

Select two emotions to compute their complexity score
Pipeline

From raw text to structured insight

Send any text to /v1/analyze and get back detected emotions, named dyads, complexity scores, and empathic levels.

01

Text Input

Submit any text up to 5,000 characters in any language supported by the LLM.

02

LLM Detection

Claude or GPT identifies emotions present with intensity and textual evidence.

03

Plutchik Mapping

Each detection maps to the 8 base emotions with PAD coordinates.

04

C_dyade Score

Weighted PAD distance yields a complexity score from 0 to 198 with an empathic level.

Integration

Three lines to emotional intelligence

Drop-in REST API

Standard JSON over HTTP. No SDK required. Works from any language, any platform. Auth via Bearer token.

RESTProtocol
JSONFormat
< 5msCompute

      
See Pricing · Free tier included

From $0/mo. Scale when you're ready.

Pricing

Start free, scale when ready

No credit card required. Upgrade anytime.

Free
$0/mo

For prototyping and exploration. Full model access with daily limits.

  • 100 compute / day
  • 10 analyze / day
  • All 6 endpoints
  • Community support
Get Started

No credit card • No expiration

Enterprise
Custom

Unlimited volume, dedicated infrastructure, SLA guarantees, and custom integrations.

  • Unlimited requests
  • 99.9% SLA
  • Dedicated support
  • Custom model tuning
Contact Sales
FAQ

Common questions

How is this different from sentiment analysis?

Sentiment analysis tells you positive or negative. Fossier tells you how complex the emotional state is. Joy + Fear (guilt) and Joy + Trust (love) both read as "mixed" in sentiment analysis. Fossier gives them distinct scores (148 vs 27), named dyads, and empathic levels.

What languages are supported?

The /v1/compute endpoint is language-agnostic: it takes emotion labels directly. The /v1/analyze endpoint processes raw text via LLM (Claude or GPT), so it works in any language your LLM supports.

What's the latency?

< 5ms for /v1/compute (pure math, no ML inference). ~1-2s for /v1/analyze (includes LLM call for emotion detection). The compute endpoint is ideal for real-time applications like games and chatbots.

Why Plutchik's model?

Plutchik's wheel is the most empirically validated discrete emotion taxonomy. It provides 8 base emotions with clear oppositions and graded intensities. Our PAD axis weights (V: 1.30, A: 0.74, D: 0.96) are cross-validated on 28,218 texts from two independent datasets (XED & SemEval).

Can I self-host?

The Enterprise plan includes self-hosted deployment options with dedicated support and custom model tuning. The compute engine is a single Python module with zero external dependencies. Contact us for details.

What if I only need the compute, not the LLM analysis?

The /v1/compute endpoint takes two emotion labels and two intensity values. No LLM needed. It's pure math at < 5ms. Use your own emotion detection pipeline and let Fossier handle the complexity scoring.

Why not just use GPT sentiment analysis?

GPT gives you a label: “mixed emotions.” Fossier gives you a score (C=148), a named dyad (guilt = joy + fear), and an empathic level (creation). That granularity lets your system decide how to respond, not just that there are emotions. Plus, /v1/compute runs in < 5ms with zero LLM cost.