18 evidence rows from five source hosts: nine competitor profiles across the AI-inference market and nine analyst forecast scenarios for Nunchux AI (bear, base, bull for 2027, 2028 and 2030), all in US dollars. As of 2026-10-11 Nunchux disclosed no revenue, customer count, funding or valuation, so every Nunchux number here is an analyst scenario anchored to its September 2026 launch — not company guidance. Several competitor figures are third-party estimates or media reports rather than audited numbers.
Each bar spans the low–high analyst range, log scale, USD millions. Highly uncertain scenarios, not company guidance; base-case valuations assume roughly 5–15x forward revenue. Hover a bar for exact figures.
Latest reported or estimated annualized revenue, log scale, USD millions. Figures are media reports and third-party estimates (Sacra and similar), not audited. fal.ai — also a media-first many-model inference API with a proprietary serving stack — is the closest comparable: ~$400M annualized revenue, $4.5B last confirmed valuation, 2.5M developers.
| Company | Type | Scope / basis | Key figures | Source |
|---|---|---|---|---|
| Baseten | Competitor | Builds infrastructure for AI inference, offering dedicated compute for open-source, fine-tuned, and custom models, alongside model APIs for | Rev: Approximately $600M annualized revenue in Q1 2026, up from $200M at the end of 2025 · Val: Up to $13 billion post-money valuation, June 2026 (two tranches priced at $11 billion and $13 billion) | pomegra.io |
| Fal.ai | Competitor | Serverless inference cloud built specifically for generative media: a curated gallery of 1,000+ production-ready image, video, audio, and 3D | Rev: ~$400M annualized as of early 2026, per Sacra (unconfirmed); reported as having doubled since October 2025 · Val: $4.5B post-money valuation at the $140M Series D in December 2025; reported (not confirmed) talks in March 202 | rywalker.com |
| Fireworks AI | Competitor | LLM-centric speed-focused inference; overlaps with fal on open-weight serving but not on the video/audio gallery | Rev: $800 million ARR, in the same period (2026) · Val: $15 billion valuation, raised at in the same period (2026) | rywalker.com |
| Groq | Competitor | Custom LPU silicon rebuilt as an inference cloud after NVIDIA's ~$20B tech license | Rev: not disclosed · Val: not disclosed | ai2.work |
| Modal Labs | Competitor | General serverless GPU compute; fal's gallery, licensed closed models, and media-tuned engine sit a layer above raw compute | Rev: ~$30M ARR (date not stated) · Val: $4.65B | rywalker.com |
| Replicate | Competitor | Closest rival to fal, now Cloudflare-owned; a breadth play with 50,000+ community models and one-line deployment of arbitrary models, aligne | Rev: ~$50M+ ARR (late 2024/early 2025) · Val: $350M (July 2023, last known) | rywalker.com |
| RunPod | Competitor | General serverless GPU compute; fal's gallery, licensed closed models, and media-tuned engine sit a layer above raw compute | Rev: $240M as of its Series A in June 2026 · Val: $1B as of its Series A in June 2026 | rywalker.com |
| Runware | Competitor | Competitor in the space of AI generation APIs for images and videos. | Rev: Approximately $1M in annual revenue (per the page's FAQ, titled [2026]); the page's stats block shows $0.7M es · Val: not disclosed | grokipedia.com |
| Together AI | Competitor | Full-stack AI-native cloud offering inference, training, and GPU clusters, LLM-first; fal is media-first and does not offer training infrast | Rev: ~$1B annualized revenue · Val: $8.3 billion valuation as of the $800 million Series C on July 1, 2026 | rywalker.com |
| Nunchux AI | Forecast 2027 — Base | Analyst scenario, not company guidance; anchored to the September 2026 public launch | Revenue $3M–$8M · Valuation $40M–$120M · 45% probability | nunchux.ai |
| Nunchux AI | Forecast 2027 — Bear | Analyst scenario, not company guidance; anchored to the September 2026 public launch | Revenue $0.5M–$2M · Valuation $15M–$50M · 40% probability | nunchux.ai |
| Nunchux AI | Forecast 2027 — Bull | Analyst scenario, not company guidance; anchored to the September 2026 public launch | Revenue $10M–$25M · Valuation $150M–$400M · 15% probability | nunchux.ai |
| Nunchux AI | Forecast 2028 — Base | Analyst scenario, not company guidance; anchored to the September 2026 public launch | Revenue $12M–$30M · Valuation $120M–$400M · 45% probability | nunchux.ai |
| Nunchux AI | Forecast 2028 — Bear | Analyst scenario, not company guidance; anchored to the September 2026 public launch | Revenue $1M–$5M · Valuation $10M–$70M · 40% probability | nunchux.ai |
| Nunchux AI | Forecast 2028 — Bull | Analyst scenario, not company guidance; anchored to the September 2026 public launch | Revenue $50M–$120M · Valuation $700M–$2B · 15% probability | nunchux.ai |
| Nunchux AI | Forecast 2030 — Base | Analyst scenario, not company guidance; anchored to the September 2026 public launch | Revenue $60M–$150M · Valuation $600M–$2.2B · 45% probability | nunchux.ai |
| Nunchux AI | Forecast 2030 — Bear | Analyst scenario, not company guidance; anchored to the September 2026 public launch | Revenue $3M–$15M · Valuation $10M–$120M · 40% probability | nunchux.ai |
| Nunchux AI | Forecast 2030 — Bull | Analyst scenario, not company guidance; anchored to the September 2026 public launch | Revenue $250M–$600M · Valuation $4B–$10B · 15% probability | nunchux.ai |
Method: 18 rows gathered as of 2026-10-11 from five source hosts — nine competitor profiles (latest funding, valuation, disclosed or estimated revenue, traction; several are media reports or third-party estimates, not audited) and nine analyst-built Nunchux scenarios (bear/base/bull for 2027/2028/2030, revenue and valuation ranges in USD millions with assigned probabilities of 40/45/15%). Nunchux forecast rows share one official Nunchux source because the numbers are original analyst scenarios anchored to public launch facts. Long scope descriptions are truncated in the table for space.