Stop picking models. Start routing
Model Navigator reads each prompt and routes it to the best-fit model for that specific task, scored on GMI's own, in-house-trained benchmark. The result: better quality where it matters, better price where it doesn't, matched to every scenario automatically
ROUTED BY GMI'S OWN TRAINED MODEL
EVERY PROMPT PARSED & TASK-MATCHED
SCORED ON GMI'S IN-HOUSE-TRAINED BENCHMARK
15
Benchmarked models in the routing pool
8
Task categories
4303
Reference prompts
Model recommendation & routing, free for a limited time. Point every prompt at its best-fit model, on us
No model wins every task. So don't bet on one
Model Navigator is built on GMI's own benchmark across real-world task categories. Switch a category and watch the best model change with the workload, the whole reason routing beats a fixed choice
A selection of popular models we currently disclose, more are benchmarked internally but not published for now. Each model carries a quality score and a cost score per task
Quality scores come from GMI's internal benchmark bank; cost scores from an internal auto-eval. Figures are directional and shown for illustration, raw sources and scoring internals aren't disclosed, and cost does not imply a fixed price, which depends on your prompt
Read the prompt. Route the model
Routing runs on GMI's own machine-learning model, trained in-house and benchmark-backed. It reads the workload before it picks the model, and it's included at no extra cost
It starts by knowing what you're asking
Before a single token is generated, GMI's routing layer parses each prompt and matches it against our prompt database, millions of reference prompts mapped to real task types. That match is what lets routing pick the right model for the job, not just any model
4303
Reference prompts in our database
8
Task types mapped
Parse
Match the prompt to our prompt database and read intent
Detect
Identify the task category behind the request
Filter
Apply model pool, scope, price tier, and context length
Evaluate
Score eligible models on quality, cost, and reliability
Route
Send the prompt to the best-fit model and return its output directly, no extra calls or steps on your side
THE 8 TASK CATEGORIES NAVIGATOR DETECTS
Coding
Refactors, tests, fixes
Agent / Tool Use
Multi-step tool calls
Reasoning
Multi-step logic
Language
Multilingual & linguistic tasks
Instruction Following
Strict output formats
Knowledge
Factual Q&A
Data Analysis
Tables, trends
Math
Proofs, computation
No routing fees. No markup. Parsing and routing run on GMI's own model and add nothing on top, the intelligence that picks your model is free
Cost — lower-cost models, quality kept acceptable
Balanced — quality, price, and reliability in balance
Quality — strongest models for high-value tasks
Allowed Model Pool — route only within models your org owner approves
Auto Mode — turn automatic routing on or off, workspace-wide
One model, every prompt. That's the tax you pay
Frontier models are brilliant, and expensive. Send a one-line prompt to your biggest model and you burn budget. Send a hard reasoning task to a weak one and you ship worse answers. Choosing by hand, request by request, doesn't scale
Fixed single model
- Every prompt hits the same model
- You overpay on the easy requests
- You underperform on the hard ones
- Quality vs. spend is a manual guess
Model Navigator
ROUTED- Reads the task behind every prompt
- Routes each request to its best-fit model
- Saves your strongest models for hard tasks
- Uses cost-efficient models where quality holds
Better answers. Lower cost. Proven
In experiments run by the GMI team, routing each task to its best-fit model beats relying on one fixed frontier model for every prompt
+2.4 pts
Higher quality than only using GPT-5.5 — Quality Mode
−28%
Lower cost per task than GPT-5.5 — Quality Mode
−84%
Lower cost per task than Opus 4.8 — Balanced Mode
Based on GMI's internal benchmark evaluation across 4,000+ evaluation tasks and 8 task categories, comparing 10 GMI-served models head-to-head. Model Navigator routes across the full 115+ LLM catalog. Quality is a benchmark score; cost is shown relative to fixed-model baselines
Quality vs cost
UP-LEFT IS BETTER ↖Always GPT-5.5
82.5% · 1.00× cost
Always Opus 4.8
81.3% · 1.38× cost
Navigator · Quality
84.9% · 0.72× cost
Navigator · Balanced
81.3% · 0.22× cost
COST PER TASK VS only using frontier model
Benchmark results are based on GMI internal evaluation and may vary by workload, prompt length, allowed model pool, and routing settings
15 benchmarked models. One routing layer
Model Navigator routes across GMI's full LLM catalog, advanced and popular open-source and closed-source models alike. Define an allowed model pool and route only within what your organization approves
ALLOWED MODEL POOL
Keep routing within approved models
MODEL SCOPE
Open-source, closed-source, or all models
PRICE TIER
Control the eligible model cost range
AUTO MODE
Enable or disable automatic routing
15
MODELS IN THE V1 ROUTING POOL
Open + Closed-source
MODEL FAMILIES
15
MODELS IN THE V1 ROUTING POOL
Open + Closed-source
MODEL FAMILIES
One endpoint. Production routing
Send requests through the dedicated Model Navigator API. It applies your workspace routing settings, picks the best-fit model, and returns the result with full routing metadata
Built for production workflows
- Dedicated routing endpoint
- Applies your workspace routing settings
- Selected model returned in routing metadata
- Authenticated with your GMI API key
Route in production. Reliably
Routing quality is only half the story. Model Navigator runs on GMI's own inference infrastructure, with the controls and failover production teams expect
Same-network inference
Models run on GMI's own infrastructure, keeping routing and generation on one low-latency network
Automatic retry & failover
If a model fails, times out, or is rate-limited, Navigator instantly reroutes to a healthy backup, so model outages rarely surface to your users
Governance built in
Admin-managed allowed model pool, model scope, and price tier, routing stays inside your policy
Workspace-level control
Routing settings and Auto Mode are configured per workspace and persist across sessions
99.9%+ availability target
Instant failover across a deep model pool means a single model going down rarely reaches your workload, effective downtime stays minimal, backed by GMI Cloud's production platform and global data centers
Full production metrics — routing latency, throughput, and error rate under load — are being benchmarked and will be published here
Stop overpaying. Start routing
Cut unnecessary model spend while keeping benchmark-backed quality across every real AI workload. Model recommendation and routing are free, for a limited time