SAVRN
Search Contact SAVRN

SAVRN Model Hub · Comparisons

chandra-ocr-2 vs medgemma-4b-it

Chandra-ocr-2 has 5.3B parameters and medgemma-4b-it has 4.3B parameters; chandra-ocr-2 is released under Open RAIL License and medgemma-4b-it under other; at 16-bit, chandra-ocr-2 needs about 12.7 GB (1x MI300X from $1.85 an hour) and medgemma-4b-it about 10.3 GB (1x MI300X from $1.85 an hour).

Published metadata for 2 models, each read from its own repository.
Field chandra-ocr-2
datalab-to/chandra-ocr-2
medgemma-4b-it
google/medgemma-4b-it
Publisher Datalab Google
Task Image and text to text Image and text to text
Modality Image and text Image and text
Parameters, as reported 5.3B parameters 4.3B parameters
Architecture Qwen3_5ForConditionalGeneration Gemma3ForConditionalGeneration
Library transformers transformers
Context length 262,144 tokens Not stated
Repository size 10.6 GB 8.6 GB
Artifact formats safetensors safetensors
License openrail other
Access Open weights, no gate Access requested at publisher
Memory at 16-bit (weights and margin) 12.7 GB 10.3 GB
Cheapest GPUs at 16-bit, per hour 1x MI300X, $1.85 1x MI300X, $1.85
Memory at 4-bit (weights and margin) 3.2 GB 2.6 GB
Cheapest GPUs at 4-bit, per hour 1x MI300X, $1.85 1x MI300X, $1.85
Revision viewed af93b47dba1b 290cda5eeccb
Downloads reported by the hub 2.8M 1.1M
Last observed 2026-09-19 2026-09-19

An evaluation row appears only where at least two of these models report the same benchmark with the same stated configuration, metric, unit and setup. Different evaluators stay named in each cell. Values are shown as reported: no unit conversion, no ranking.

Other Reported Results

These results are listed for each model on its own, because the conditions needed to compare them are not stated or do not match. Two results that leave a condition blank are not assumed to share it.

chandra-ocr-2

BenchmarkConditionsResultReported byRevisionDate
Delores-Lin/MDPBench Task arMetric arComparison conditions not established 78.2 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task deMetric deComparison conditions not established 86.6 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task digitalMetric digitalComparison conditions not established 87.8 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task enMetric enComparison conditions not established 86.5 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task esMetric esComparison conditions not established 69.7 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task frMetric frComparison conditions not established 70.3 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task hiMetric hiComparison conditions not established 81.1 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task idMetric idComparison conditions not established 84.6 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task itMetric itComparison conditions not established 87.4 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task jpMetric jpComparison conditions not established 68.8 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task koMetric koComparison conditions not established 80.3 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task latinMetric latinComparison conditions not established 82.7 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task nlMetric nlComparison conditions not established 82.7 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task non_latinMetric non_latinComparison conditions not established 76.4 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task overallMetric overallComparison conditions not established 79.7 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task photographedMetric photographedComparison conditions not established 77.1 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task privateMetric privateComparison conditions not established 82.4 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task ptMetric ptComparison conditions not established 90.7 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task ruMetric ruComparison conditions not established 74 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task thMetric thComparison conditions not established 78.5 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task viMetric viComparison conditions not established 85.6 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task zhMetric zhComparison conditions not established 73.8 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
Delores-Lin/MDPBench Task zh_tMetric zh_tComparison conditions not established 76.3 MDPBench leaderboard
Reported by a third party
Evaluated revision not stated 2026-07-03
allenai/olmOCR-bench Task arxiv_mathMetric arxiv_mathComparison conditions not established 86.9 Chandra OCR 2 Model Card
Reported by a third party
Evaluated revision not stated 2026-06-26
allenai/olmOCR-bench Task baselineMetric baselineComparison conditions not established 99.9 Chandra OCR 2 Model Card
Reported by a third party
Evaluated revision not stated 2026-06-26
allenai/olmOCR-bench Task headers_footersMetric headers_footersComparison conditions not established 91.4 Chandra OCR 2 Model Card
Reported by a third party
Evaluated revision not stated 2026-06-26
allenai/olmOCR-bench Task long_tiny_textMetric long_tiny_textComparison conditions not established 93.7 Chandra OCR 2 Model Card
Reported by a third party
Evaluated revision not stated 2026-06-26
allenai/olmOCR-bench Task multi_columnMetric multi_columnComparison conditions not established 82.1 Chandra OCR 2 Model Card
Reported by a third party
Evaluated revision not stated 2026-06-26
allenai/olmOCR-bench Task old_scansMetric old_scansComparison conditions not established 51.1 Chandra OCR 2 Model Card
Reported by a third party
Evaluated revision not stated 2026-06-26
allenai/olmOCR-bench Task old_scans_mathMetric old_scans_mathComparison conditions not established 89.1 Chandra OCR 2 Model Card
Reported by a third party
Evaluated revision not stated 2026-06-26
allenai/olmOCR-bench Task overallMetric overallComparison conditions not established 85.8 Chandra OCR 2 Model Card
Reported by a third party
Evaluated revision not stated 2026-06-26
allenai/olmOCR-bench Task table_testsMetric table_testsComparison conditions not established 92.1 Chandra OCR 2 Model Card
Reported by a third party
Evaluated revision not stated 2026-06-26
llamaindex/ParseBench Task chartMetric chartSetup Pipeline name: chandra2_vllmComparison conditions not established 65.1 ParseBench
Reported by a third party
Evaluated revision not stated 2026-04-14
llamaindex/ParseBench Task layoutMetric layoutSetup Pipeline name: chandra2_vllmComparison conditions not established 51.2 ParseBench
Reported by a third party
Evaluated revision not stated 2026-04-14
llamaindex/ParseBench Task meanMetric meanSetup Pipeline name: chandra2_vllmComparison conditions not established 70.1 ParseBench
Reported by a third party
Evaluated revision not stated 2026-04-14
llamaindex/ParseBench Task tableMetric tableSetup Pipeline name: chandra2_vllmComparison conditions not established 89.2 ParseBench
Reported by a third party
Evaluated revision not stated 2026-04-14
llamaindex/ParseBench Task text_contentMetric text_contentSetup Pipeline name: chandra2_vllmComparison conditions not established 83.7 ParseBench
Reported by a third party
Evaluated revision not stated 2026-04-14
llamaindex/ParseBench Task text_formattingMetric text_formattingSetup Pipeline name: chandra2_vllmComparison conditions not established 61.4 ParseBench
Reported by a third party
Evaluated revision not stated 2026-04-14

SAVRN's Notes on chandra-ocr-2

Chandra 2 turns images and PDFs into markdown, HTML or JSON while keeping the page layout, which is the part most OCR tools lose. Datalab built it on a Qwen 3.5 vision base at 5.3B parameters, and 2.8 million downloads a month make it one of the most used document models in our catalog.

It fits on a single mid-range GPU at 16-bit, and a 262,144-token context handles long multipage output. The OpenRAIL license allows commercial use subject to its use restrictions so read them before building a paid document product on it. For a document-heavy business this is a model worth testing first.

SAVRN's Notes on medgemma-4b-it

Medical image and text comprehension is what Google trained this for, and the 4B instruction-tuned build is the smallest of three MedGemma variants; the other two are 27B. At 16-bit the weights are 8.6 GB and the run needs 10.3 GB, a sliver of the 192 GB on one MI300X at $1.85 an hour on-demand. At 8-bit the need drops to 5.2 GB, at 4-bit to 2.6 GB. Fit is not the constraint; sessions per card is.

The license reads "other" and access is gated, so a deployment begins with Google's Health AI Developer Foundations terms rather than a standard open license, and those terms come before the files do. Confirm you are taking the build derived from medgemma-4b-pt, and check its context length against your prompts, since our record carries no figure. No host prices are on the Index yet, so your own facility is the priced path.

Questions

Which is larger, chandra-ocr-2 or medgemma-4b-it?

chandra-ocr-2 (5.3B parameters) is larger than medgemma-4b-it (4.3B parameters), by the parameter counts their publishers report.

Which is cheaper to run, chandra-ocr-2 or medgemma-4b-it?

At 4-bit, chandra-ocr-2 fits on 1x MI300X from $1.85 an hour and medgemma-4b-it on 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use chandra-ocr-2 commercially?

Yes, with conditions. chandra-ocr-2 is released under Open RAIL License. Open RAIL licenses permit use, including commercial use, subject to the use-based restrictions listed in the license, which must be passed on to anyone who receives the model or a derivative.

Related Comparisons