ThinEmbed
Compressed query encoders for embedding models that run in the browser
- Upvotes
- 0
- DR
- 0 no change since last check
- Followers
- 0
Query encoders compressed for efficient embedding search with unchanged document indices. Runs in the browser on WebGPU, supporting multiple base models like
About ThinEmbed
ThinEmbed provides compressed query encoders for embedding models optimized for browser deployment. The compression approach maintains the original document index unchanged while compressing only the query encoder side, enabling fast semantic search in web environments. The encoders are available in multiple compression levels (from 1.83 to 3.57 bits per weight) and support various base models including Qwen3-Embedding-0.6B and microsoft/harrier-oss-v1-0.6b. The system can run on WebGPU with shader-f16 support in browsers like Chrome and Edge.
The project includes pre-compiled files, npm and PyPI packages for measurement and deployment, and comprehensive benchmarking results on multiple datasets including SciFact, NFCorpus, ArguAna, SciDocs, and Czech WebFAQ-cs. Quality is measured against the full-precision model's performance, with results showing nDCG@10 retention between 94% and 101.3% of the original model depending on the compression level and dataset.
- open-source
- browser
- semantic-search
- embeddings
- webgpu
- quantization
Features worth knowing
- Runs in the browser with WebGPU support (Chrome, Edge)
- Multiple compression levels from 1.83 to 3.57 bits per weight
- Preserves unchanged document embedding index
- Available as npm package and PyPI module
- Comprehensive benchmarking on SciFact, NFCorpus, ArguAna, SciDocs, and WebFAQ-cs
- Pre-registered measurements with paired bootstrap intervals
Pricing
Pricing model: Open source
Follow ThinEmbed and hear about their milestones, new launches and demo videos by email. Stop any time.
ThinEmbed metrics
| Metric | Value |
|---|---|
| Domain Rating | 0 no change since last check |
| Trust Flow | 0 no change since last check |
| Citation Flow | 0 no change since last check |
| Referring domains | 0 no change since last check |
| External backlinks | 0 no change since last check |
Ahrefs Domain Rating scanned 6 Nov 2026, updated 3 Nov 2026. Majestic scanned 3 Nov 2026, updated 3 Nov 2026. Link stats via Majestic. Trust Flow, Citation Flow and Topical Trust Flow are trademarks of Majestic-12 Ltd.
ThinEmbed domain facts
Reveal the domain facts
| Domain age | 7 days |
|---|---|
| Registered | |
| Expires | |
| Registrar | NameCheap, Inc. |
| DNSSEC | Not signed |
ThinEmbed launch results
| Measure | At launch, then latest |
|---|---|
| Domain Rating (Ahrefs) | 0 to 0 no change since launch 3 Nov 2026, then 5 Nov 2026 |
| Trust Flow (Majestic) | 0 first measured 3 Nov 2026 |
| Referring domains (Majestic) | 0 first measured 3 Nov 2026 |
Day 2 since launch. Next receipt: day 30, 3 Dec 2026. The figures report a change; they do not say what caused it, and many things can. No figure here is a guess.
Launching something soon?
Free tips from Best Digital Launch for getting ready, launch day, the week after and what to do next.
- Set up
- On launch day
- The days after
- Beyond launch
Compare and explore
FAQ
- Does compressing the query encoder require re-encoding documents?
- No. The compression only affects the query encoder. The document index stays exactly as it is, so existing indices do not need to be re-encoded or modified.
- What browser runtime does ThinEmbed require?
- ThinEmbed requires WebGPU with shader-f16 support and runs in Chrome and Edge. There is no CPU fallback available; WebGPU is mandatory.
Leave a review for ThinEmbed
Nobody has reviewed ThinEmbed yet. Start things off.
A review is always free and can never be paid for. Go to the reviews and discussion
Reviews and chat · Product news · Send it on · Run this company? Claim it here · Report it
ThinEmbed has a free listing. The owner can unlock the full profile with a paid launch for $5:
- Founder story (locked)
- Milestones (locked)
- Demo video (locked)
- Dofollow website link (locked)