GILD Intelligence
Inference hosts
2 of the 88 AI products we track, each with the day we last checked it.
- Fireworks AIHosted inference for open models: shared Serverless billed per token, on-demand GPUs billed per GPU second, and fine-tuning.Category: Inference HostingFireworks AIUpdated: Nov 1, 2026Practitioner sentiment: −17 Mixed
- Together AIInference host for open models: a serverless per-token API, reserved capacity, fine-tuning and GPU clusters.Category: Inference HostingTogether AIUpdated: Nov 1, 2026Practitioner sentiment: −49 Negative
Where each product sits
Every product we track, placed by the stage of work it mainly serves. Building software with agents holds the most.
Agentic Software Development
1Plan
32Build
4Validate
0Ship
None yet
0Operate
None yet
AI Application Platform
11Model
6Build
3Data
5Evaluate
2Operate
AI Knowledge Work
8Assist
3Find
4Create
4Analyze
5Automate
What each stage holds
Agentic Software Development
- Plan
- Work planning
- Build
- Codebase context, Coding, Agent runtime, Tool execution
- Validate
- Debugging, Testing, Code review, Application security
- Ship
- Source control, CI/CD
- Operate
- Observability
AI Application Platform
- Model
- Model APIs, Model gateway
- Build
- Orchestration, Tool use
- Data
- Retrieval
- Evaluate
- Evaluation
- Operate
- Observability, Cloud deployment, Guardrails
AI Knowledge Work
- Assist
- AI assistant
- Find
- Web research, Enterprise search
- Create
- Documents, Meeting notes
- Analyze
- Data analysis
- Automate
- Automation