Udgam 384M Chat has 0.384 billion total parameters. The selected Q4_K_M GGUF file is 0.238 GB.
Data snapshot: . 1 published test in this catalog.
Model file and details
Provider
UdgamLabs
Total parameters
384,389,332
Exact file size
238,124,448 bytes
Quantization / precision
Q4_K_M
Weight repository
UdgamLabs/Udgam-384M-Chat-GGUF
Original model repository
UdgamLabs/Udgam-384M-Chat
Weight revision
387d69c11923f6ac06ff4b7adf467b5bd22bf036
Model revision
bd39c0291df708857172c2c652d988154b3238c7
Runtime notes
Confirm model architecture and quantization support in your runner.
Recorded release date
2026-10-09
Date evidence
UdgamLabs created the Apache-2.0 source checkpoint and official GGUF repository on 9 October 2026. The selected 238,124,448-byte Q4_K_M artifact is the publisher's low-memory recommendation and is the exact quantization used for its documented function-calling and llama.cpp measurements.
These scores were compiled from the linked sources. These published scores were not independently tested here. Different prompts, precision, thinking budgets and tools can change results.
BFCL v3 simple · 150
62.7 / 100
Publisher-run accuracy on 150 simple function-calling cases. This is a documented subset, not the complete BFCL v3 leaderboard suite.
Benchmark efficiency: 263.308 published score points per GB of selected weights. This ratio is not an inference-speed measure; check the test settings to see whether the score covers the selected quantization.
Test settings: Publisher-run automatic checking on 150 BFCL v3 simple cases using the model-native template, llama.cpp on a Mac, greedy decoding and thinking disabled. The selected Q4_K_M artifact scored 62.7%; the same card separately reports 64.0% on 100 multiple, 50.0% on 50 parallel and 65.3% on 150 irrelevance cases.
Publisher-run and not independently reproduced. The 150-case simple set is a subset, not the complete BFCL v3 leaderboard suite. UdgamLabs describes this as an on-device tool-calling model, not a knowledgeable general assistant or safety boundary; its card warns about factual mistakes, weak exact-call and instruction-following behavior, and recommends allowlisted tools, argument validation and confirmation for risky actions. The 238,124,448-byte weight file excludes runtime and context memory. A publisher llama.cpp run observed 323 MB RAM and 14,191/444 prompt/generation tokens per second on an Apple M6 Mac mini, but the machine's memory configuration was not stated, so those measurements do not establish performance or compatibility for a purchasable configuration. Apache-2.0 license.
Context and version evidence: The repository and file name identify the selected model. Revision hashes are shown above when recorded. Tested context length, full hardware configuration and run date are not consistently recorded in this snapshot. Only settings explicitly stated in each benchmark record are known.
The 0.238 GB figure covers model weights in this file. It excludes the context cache, runtime, operating system and any separate multimodal files. It is not a measured RAM requirement or a token-generation speed.
These complete configurations pass the initial weight-size check for the selected 0.24 GB file. Allow extra memory for context and software. Check the model’s runtime notes and your inference software before buying.
Listing evidence: 2026-09-25. Amazon.com base price visible in an owner-supplied screenshot reviewed 25 Sep 2026; capture time not supplied. Dated reference, not a live quote. Excludes card signup offers, coupons, tax and shipping.
Listing evidence: 2026-10-03. Apple US launch price announced 22 September 2026: USD 1,699. Dated manufacturer reference, not an Amazon offer; tax and shipping may differ.
Listing evidence: 2026-09-24. GMKtec, EVO-X2 / 64GB / 1TB, recorded 22 Sep 2026; exact ASIN capacity identified 24 Sep. This is a dated reference, not a current Amazon price; tax, shipping and bundle contents may differ.
Memory matches are estimates, not tested compatibility or speed. As an Amazon Associate I earn from qualifying purchases.
Confirm model architecture and quantization support in your runner. The selected weights use 0.238 GB; this excludes context cache, runtime buffers and other programs.
These are the first listed tiers that pass our reserve policy, not measured minimum requirements. System RAM and VRAM remain separate. Use the memory tool to change the reserve for your context and runner.