The State of Translation Automation 2026 · Languages

English to Macedonian AI translation quality, cost and speed

What it takes to automate translation from English into Macedonian against 10 requirements in three levels: the solutions that meet them, what they cost and how fast they are, from the 2026 benchmark of 57 AI models and MT engines.

How good is AI translation from English to Macedonian?

Key findings on English to Macedonian translation

  1. 1Metrics vs LQATwo of the four solutions that meet every Standard requirement in Macedonian, on OpenAI GPT-5.5 and Anthropic Claude Fable 5, made the shortlist through MetricX alone. COMET-Kiwi ranks them 11th and 12th.See the data →
  2. 2Real-timeEight of the nine real-time (<2 sec) models in Macedonian are MT engines or LLM-based MT engines. None of the three on the shortlist meet general translation quality, and translations from the nearest, Translated Lara 2, have 24.7 weighted errors per 1,000 words.See the data →

Find English to Macedonian translation solutions for your requirements

Pick a quality level, or your own requirements, bars and constraints, to see the shortlisted solutions that meet them in Macedonian, with a column for each requirement you select. Under Optimize for, choose which of them comes first: the one with the fewest total errors, the lowest cost or the shortest time. Click a column heading to sort by it.

The shortlisted solutions on 11 models, with their Intento LQA results. Total errors per 1,000 words add the weighted general translation quality errors (minor 1, major 5, critical 10) to the missed glossary terms, sentences in the wrong register, tag errors and sentences that fail gender neutrality, each counted once. Formatting counts as met or missed and does not add to the total errors. The models that did not make the shortlist have automatic scores only: the shortlist. Gender neutrality: the Advanced requirements count the two best solutions of the language, with the gender-neutrality agent where they need it. The results of the other solutions are shown for information.

How we measure it: 11-01 · 11-02

Trade-offs in English to Macedonian translation

Trade-off

Four one-step solutions meet every Standard requirement in Macedonian. The one on OpenAI GPT-5.6 Terra has the fewest total errors, 14.9 per 1,000 words, the lowest cost and the shortest time, so Optimize for selects it for quality, cost and latency alike. Among the other three, the solution on Google Gemini 3.1 Pro Preview costs $5.09 per million characters against $12.01 for the one on OpenAI GPT-5.5, but has 22.5 total errors per 1,000 words against 16.2 and takes 97 seconds per 1,000 words against 73.

Constraints

Under the Standard requirements, a cost cap or a time limit below those of the solution on OpenAI GPT-5.6 Terra, $4.84 per million characters and 32 seconds per 1,000 words, removes all four solutions. A limit on the steps keeps all four, since each takes one step. Under the Advanced requirements, any limit on the solution complexity removes the only solution, because it works with the gender-neutrality agent. So does a cost cap below $12.57 per million characters or a time limit below 64 seconds per 1,000 words.

Agent effect

The gender-neutrality agent takes the solution on OpenAI GPT-5.5 from $12.01 to $12.57 per million characters, from 73 to 64 seconds per 1,000 words and from 16.2 to 12.2 total errors per 1,000 words. On OpenAI GPT-5.6 Terra, the agent comes in a three-step solution (translation + LQA + repair), so the change combines two extra steps with the agent. That solution has 2.9 total errors per 1,000 words against 14.9 for the one-step solution, at $976 against $4.84 per million characters and 160 against 32 seconds per 1,000 words.

English to Macedonian translation quality levels

Optimize for
Basic requirements★★★★

Tier 1: simple solution + backup

Solutions that meet them
4 (4 one-step)
Simplest solution
one-step
Fewest errors

OpenAI GPT-5.6 Terra
one-step LLM adjusted for Standard requirements

Total errors per 1,000 words
14.9
Cost per million characters
$4.84
Seconds per 1,000 words
32
Lowest cost

OpenAI GPT-5.6 Terra
one-step LLM adjusted for Standard requirements

Total errors per 1,000 words
14.9
Cost per million characters
$4.84
Seconds per 1,000 words
32
Shortest time

OpenAI GPT-5.6 Terra
one-step LLM adjusted for Standard requirements

Total errors per 1,000 words
14.9
Cost per million characters
$4.84
Seconds per 1,000 words
32
Standard requirements★★★★

Tier 1: simple solution + backup

Solutions that meet them
4 (4 one-step)
Simplest solution
one-step
Fewest errors

OpenAI GPT-5.6 Terra
one-step LLM adjusted for Standard requirements

Total errors per 1,000 words
14.9
Cost per million characters
$4.84
Seconds per 1,000 words
32
Lowest cost

OpenAI GPT-5.6 Terra
one-step LLM adjusted for Standard requirements

Total errors per 1,000 words
14.9
Cost per million characters
$4.84
Seconds per 1,000 words
32
Shortest time

OpenAI GPT-5.6 Terra
one-step LLM adjusted for Standard requirements

Total errors per 1,000 words
14.9
Cost per million characters
$4.84
Seconds per 1,000 words
32
Advanced requirements☆

Tier 3: one solution

Solutions that meet them
1 (0 one-step)
Simplest solution
with the gender-neutrality agent
Fewest errors

OpenAI GPT-5.5
one-step with the gender-neutrality agent

Total errors per 1,000 words
12.2
Cost per million characters
$12.57
Seconds per 1,000 words
64
Lowest cost

OpenAI GPT-5.5
one-step with the gender-neutrality agent

Total errors per 1,000 words
12.2
Cost per million characters
$12.57
Seconds per 1,000 words
64
Shortest time

OpenAI GPT-5.5
one-step with the gender-neutrality agent

Total errors per 1,000 words
12.2
Cost per million characters
$12.57
Seconds per 1,000 words
64

Optimize for selects one of the solutions that meet the level: the one with the fewest total errors, the lowest cost or the shortest time. Where no solution meets the level, the card shows the available solution with the fewest total errors and what it misses. Total errors count every requirement with countable errors.

Quality levelTierSolutionsSimplest solutionModelTotal errors per 1,000 wordsCost per million charactersSeconds per 1,000 words
BasicTier 1: simple solution + backup★×4one-stepOpenAI GPT-5.6 Terra14.9$4.8432
StandardTier 1: simple solution + backup★×4one-stepOpenAI GPT-5.6 Terra14.9$4.8432
AdvancedTier 3: one solution☆with the gender-neutrality agentOpenAI GPT-5.512.2$12.5764

One solution per level: the simplest that meets the level, the cheapest of them where several are equally simple. ★ a one-step solution, ☆ a solution with more steps or the gender-neutrality agent.

Level step-up

Optimize for selects one solution for quality, cost and latency alike: the one-step solution on OpenAI GPT-5.6 Terra under the Basic and Standard requirements, and the one on OpenAI GPT-5.5 with the gender-neutrality agent under the Advanced ones. The step to the Advanced requirements removes all four one-step solutions and takes the agent. The only solution that meets them has 12.2 total errors per 1,000 words against 14.9 for the one on OpenAI GPT-5.6 Terra, at $12.57 per million characters and 64 seconds per 1,000 words.

How we measure it: 11-01 · 11-02 · 08-09

English to Macedonian translation cost against errors

Chart data
Cost and weighted errors of the one-step solutions in Macedonian
ModelUS dollars per million charactersWeighted errors per 1,000 wordsGroup
Google Gemma 40.1542.8Misses a Standard requirement
z.ai GLM-52.141.9Misses a Standard requirement
Google Gemini 3.5 Flash3.830.7Misses a Standard requirement
OpenAI GPT-5.6 Terra4.8411.9Meets every Standard requirement
Google Gemini 3.1 Pro Preview5.0919.4Meets every Standard requirement
Anthropic Claude Sonnet 55.3248.8Misses a Standard requirement
Microsoft Translator1057.7Misses a Standard requirement
OpenAI GPT-5.512.0113.5Meets every Standard requirement
Translated Lara 224.9924.7Misses a Standard requirement
Anthropic Claude Fable 526.9517.5Meets every Standard requirement
DeepL Next Gen27.543.2Misses a Standard requirement

How we measure it: 11-01 · 10-10

English to Macedonian translation requirements

Chart data
Errors per 1,000 words of the one-step solutions in Macedonian, median, by requirement
RequirementMedian errors per 1,000 words
General translation quality30.7
Gender neutrality2.8
Tone of voice0.1
Tag handling0

The requirements with countable errors. Every solution with its result on each requirement: the explorer.

How we measure it: 09-09 · 09-10 · 09-11 · 09-12 · 09-18

AI models and MT engines for English to Macedonian translation

Chart data
Weighted errors per 1,000 words of the shortlisted one-step solutions in Macedonian
ModelWeighted errors per 1,000 words of the shortlisted one-step solutions in MacedonianClass
Microsoft Translator57.7MT engines
Anthropic Claude Sonnet 548.8LLMs
DeepL Next Gen43.2LLM-based MT engines
Google Gemma 442.8Open-weight LLMs
z.ai GLM-541.9LLMs
Google Gemini 3.5 Flash30.7LLMs
Translated Lara 224.7LLM-based MT engines
Google Gemini 3.1 Pro Preview19.4LLMs
Anthropic Claude Fable 517.5LLMs
OpenAI GPT-5.513.5LLMs
OpenAI GPT-5.6 Terra11.9LLMs
Model types

On the shortlist, translations from the eight LLMs have 11.9 to 48.8 weighted errors per 1,000 words, and the four solutions that meet general translation quality are all on LLMs. Translations from the two LLM-based MT engines, Translated Lara 2 and DeepL Next Gen, have 24.7 and 43.2 weighted errors per 1,000 words, and translations from the MT engine, Microsoft Translator, have 57.7. All eight shortlisted LLMs, the open-weight Google Gemma 4 among them, support Macedonian by our measured criteria.

How we measure it: 00-04 · 08-15 · 23-01

How we shortlisted English to Macedonian translation solutions

Chart data
Ranks of the models on the two automatic metrics, Macedonian
ModelCOMET-KiwiRank on COMET-KiwiMetricXRank on MetricXShortlistedShortlist place
Anthropic Claude Opus 4.80.80691.3310NoNot shortlisted
Anthropic Claude Sonnet 4.60.806101.3514NoNot shortlisted
OpenAI GPT-4.10.804131.3413NoNot shortlisted
xAI Grok 4.20 (non-reasoning)0.802151.3411NoNot shortlisted
DeepSeek V3.20.799161.4419NoNot shortlisted
Tilde TildeOpen LLM engine0.799171.3918NoNot shortlisted
Google Translation LLM0.799181.5724NoNot shortlisted
Mistral Large 30.797191.3615NoNot shortlisted
ModernMT Enterprise Edition0.797201.3716NoNot shortlisted
MoonshotAI Kimi K2.50.794211.4621NoNot shortlisted
Mistral Medium 3.10.791221.4922NoNot shortlisted
MiniMax M2.50.791231.5123NoNot shortlisted
Anthropic Claude Haiku 4.50.790241.5926NoNot shortlisted
Swiss AI Apertus-8B-25090.788251.6127NoNot shortlisted
Alibaba Qwen3-235B-A22B0.783261.4620NoNot shortlisted
SYSTRAN PNMT API0.782271.7228NoNot shortlisted
OpenAI gpt-oss-120b0.778281.5925NoNot shortlisted
Amazon Translate0.778291.8129NoNot shortlisted
Google Cloud Advanced Translation API0.771302.1831NoNot shortlisted
Cohere Command A Translate0.761311.8430NoNot shortlisted
Yandex Cloud Translate v20.742322.5432NoNot shortlisted
AI21 Jamba Large 1.70.693333.1434NoNot shortlisted
NiuTrans0.687342.7333NoNot shortlisted
Alibaba Machine Translation API0.624354.2335NoNot shortlisted
AppTek NMT0.457368.7236NoNot shortlisted
Google Gemini 3.1 Pro Preview0.81211.255YesTop 8 by both metrics
Translated Lara 20.81121.3917YesTop 8 by COMET-Kiwi
z.ai GLM-50.81131.299YesTop 8 by COMET-Kiwi
OpenAI GPT-5.6 Terra0.81141.141YesTop 8 by both metrics
DeepL Next Gen0.81151.277YesTop 8 by both metrics
Google Gemini 3.5 Flash0.80861.172YesTop 8 by both metrics
Anthropic Claude Sonnet 50.80871.234YesTop 8 by both metrics
Google Gemma 40.80881.298YesTop 8 by both metrics
OpenAI GPT-5.50.805111.266YesTop 8 by MetricX
Anthropic Claude Fable 50.804121.223YesTop 8 by MetricX
Microsoft Translator0.804141.3412YesReal-time place
ModelSolution complexityCOMET-Kiwi rankCOMET-KiwiMetricX rankMetricXShortlist
Google Gemini 3.1 Pro PreviewAdjusted LLM10.81251.25Top 8 by both metrics
Translated Lara 2Adjusted NMT20.811171.39Top 8 by COMET-Kiwi
z.ai GLM-5Adjusted LLM30.81191.29Top 8 by COMET-Kiwi
OpenAI GPT-5.6 TerraAdjusted LLM40.81111.14Top 8 by both metrics
DeepL Next GenOff-the-shelf NMT50.81171.27Top 8 by both metrics
Google Gemini 3.5 FlashAdjusted LLM60.80821.17Top 8 by both metrics
Anthropic Claude Sonnet 5Adjusted LLM70.80841.23Top 8 by both metrics
Google Gemma 4Adjusted LLM80.80881.29Top 8 by both metrics
Anthropic Claude Opus 4.8Adjusted LLM90.806101.33Not shortlisted
Anthropic Claude Sonnet 4.6Adjusted LLM100.806141.35Not shortlisted
OpenAI GPT-5.5Adjusted LLM110.80561.26Top 8 by MetricX
Anthropic Claude Fable 5Adjusted LLM120.80431.22Top 8 by MetricX
OpenAI GPT-4.1Adjusted LLM130.804131.34Not shortlisted
Microsoft TranslatorOff-the-shelf NMT140.804121.34Real-time place
xAI Grok 4.20 (non-reasoning)Adjusted LLM150.802111.34Not shortlisted
DeepSeek V3.2Adjusted LLM160.799191.44Not shortlisted
Tilde TildeOpen LLM engineOff-the-shelf NMT170.799181.39Not shortlisted
Google Translation LLMAdjusted NMT180.799241.57Not shortlisted
Mistral Large 3Adjusted LLM190.797151.36Not shortlisted
ModernMT Enterprise EditionOff-the-shelf NMT200.797161.37Not shortlisted
MoonshotAI Kimi K2.5Adjusted LLM210.794211.46Not shortlisted
Mistral Medium 3.1Adjusted LLM220.791221.49Not shortlisted
MiniMax M2.5Adjusted LLM230.791231.51Not shortlisted
Anthropic Claude Haiku 4.5Adjusted LLM240.790261.59Not shortlisted
Swiss AI Apertus-8B-2509Off-the-shelf LLM250.788271.61Not shortlisted
Alibaba Qwen3-235B-A22BAdjusted LLM260.783201.46Not shortlisted
SYSTRAN PNMT APIOff-the-shelf NMT270.782281.72Not shortlisted
OpenAI gpt-oss-120bAdjusted LLM280.778251.59Not shortlisted
Amazon TranslateOff-the-shelf NMT290.778291.81Not shortlisted
Google Cloud Advanced Translation APIOff-the-shelf NMT300.771312.18Not shortlisted
Cohere Command A TranslateAdjusted LLM310.761301.84Not shortlisted
Yandex Cloud Translate v2Off-the-shelf NMT320.742322.54Not shortlisted
AI21 Jamba Large 1.7Adjusted LLM330.693343.14Not shortlisted
NiuTransOff-the-shelf NMT340.687332.73Not shortlisted
Alibaba Machine Translation APIOff-the-shelf NMT350.624354.23Not shortlisted
AppTek NMTOff-the-shelf NMT360.457368.72Not shortlisted
OpenAI GPT-5.6 LunaAdjusted LLM—0.805—1.14Same model family as OpenAI GPT-5.6 Terra
OpenAI GPT-5.6 SolAdjusted LLM—0.809—1.13Same model family as OpenAI GPT-5.6 Terra
xAI Grok 4.20 (reasoning)Adjusted LLM—0.795—1.74Same model family as xAI Grok 4.20 (non-reasoning)

Every model we got a translation from, as the shortlist compared it: off the shelf, without context in the request, or adjusted, with it. A model that did not make the shortlist is shown adjusted, unless off the shelf it scores higher by 0.03 or more on COMET-Kiwi. Ranks among 36 models, with each model family counted once, and the top 8 of each metric in bold. Higher is better on COMET-Kiwi, lower on MetricX. The scores rank models within this language and are not a requirement result.

Metrics vs LQA

The metrics and Intento LQA order the shortlisted solutions differently in Macedonian. COMET-Kiwi alone brought Translated Lara 2 and z.ai GLM-5 onto the list, second and third on it, and translations from them have 24.7 and 41.9 weighted errors per 1,000 words. MetricX alone brought OpenAI GPT-5.5 and Anthropic Claude Fable 5, and the solutions on both meet every Standard requirement. Nearest the cut are Anthropic Claude Opus 4.8, 9th on COMET-Kiwi and 10th on MetricX, and Anthropic Claude Sonnet 4.6, 10th on COMET-Kiwi.

How we measure it: 00-08 · 00-09

English to Macedonian translation cost

Quality levelCheapest solution that meets every requirement of the levelCost per million characters
BasicOpenAI GPT-5.6 Terra, one-step LLM adjusted for Standard requirements$4.84
StandardOpenAI GPT-5.6 Terra, one-step LLM adjusted for Standard requirements$4.84
AdvancedOpenAI GPT-5.5, one-step with the gender-neutrality agent$12.57
Chart data
Cost of the shortlisted one-step solutions in Macedonian
ModelUS dollars per million charactersClass
Google Gemma 4$0.15Open-weight LLMs
z.ai GLM-5$2.1LLMs
Google Gemini 3.5 Flash$3.8LLMs
OpenAI GPT-5.6 Terra$4.84LLMs
Google Gemini 3.1 Pro Preview$5.09LLMs
Anthropic Claude Sonnet 5$5.32LLMs
Microsoft Translator$10MT engines
OpenAI GPT-5.5$12.01LLMs
Translated Lara 2$24.99LLM-based MT engines
Anthropic Claude Fable 5$26.95LLMs
DeepL Next Gen$27.5LLM-based MT engines
Chart data
Tokens per character of the translation into Macedonian, by model
ModelTokens per characterClass
AI21 Jamba Large 1.70.557LLMs
Mistral Medium 3.10.43LLMs
Anthropic Claude Opus 4.80.426LLMs
Mistral Large 30.426LLMs
Anthropic Claude Fable 50.425LLMs
Anthropic Claude Sonnet 50.425LLMs
Alibaba Qwen3-235B-A22B0.419Open-weight LLMs
MoonshotAI Kimi K2.50.408LLMs
Anthropic Claude Sonnet 4.60.397LLMs
Anthropic Claude Haiku 4.50.394LLMs
Cohere Command A Translate0.376LLMs
DeepSeek V3.20.372LLMs
z.ai GLM-50.364LLMs
Google Gemini 3.1 Pro Preview0.345LLMs
Google Gemini 3.5 Flash0.345LLMs
Google Gemma 40.343Open-weight LLMs
OpenAI GPT-4.10.336LLMs
xAI Grok 4.20 (non-reasoning)0.335LLMs
OpenAI GPT-5.6 Terra0.335LLMs
xAI Grok 4.20 (reasoning)0.334LLMs
OpenAI GPT-5.50.334LLMs
OpenAI GPT-5.6 Luna0.334LLMs
MiniMax M2.50.333LLMs
OpenAI GPT-5.6 Sol0.333LLMs
Cost driver

The tokenizers produce a median 0.37 tokens per character of Macedonian, 1.5 times as many as for English. Tokens alone do not set the price. The tokenizers of OpenAI GPT-5.6 Terra and OpenAI GPT-5.5 produce 0.335 and 0.334 tokens per character, yet the solutions on them cost $4.84 and $12.01 per million characters. The move from one model to the other makes most of the step to the Advanced requirements, since the gender-neutrality agent takes the solution on OpenAI GPT-5.5 only to $12.57 per million characters.

How we measure it: 10-01 · 10-03 · 05-02

English to Macedonian translation speed

Quality levelFastest solution that meets every requirement of the levelSeconds per 1,000 words
BasicOpenAI GPT-5.6 Terra, one-step LLM adjusted for Standard requirements32
StandardOpenAI GPT-5.6 Terra, one-step LLM adjusted for Standard requirements32
AdvancedOpenAI GPT-5.5, one-step with the gender-neutrality agent64
Chart data
Time to translate 1,000 words into Macedonian, by model
ModelSeconds per 1,000 wordsClass
Microsoft Translator0.2MT engines
Google Cloud Advanced Translation API0.2MT engines
ModernMT Enterprise Edition0.4MT engines
Google Translation LLM0.6LLM-based MT engines
DeepL Next Gen0.7LLM-based MT engines
SYSTRAN PNMT API1MT engines
Amazon Translate1.3MT engines
Translated Lara 21.5LLM-based MT engines
OpenAI gpt-oss-120b2.5Open-weight LLMs
Yandex Cloud Translate v22.8MT engines
Tilde TildeOpen LLM engine6.8LLM-based MT engines
Google Gemini 3.5 Flash16LLMs
Alibaba Machine Translation API16MT engines
OpenAI GPT-5.6 Luna17LLMs
xAI Grok 4.20 (non-reasoning)21LLMs
OpenAI GPT-4.122LLMs
OpenAI GPT-5.6 Terra32LLMs
Swiss AI Apertus-8B-250941Open-weight LLMs
Mistral Medium 3.145LLMs
MiniMax M2.545LLMs
Anthropic Claude Haiku 4.546LLMs
xAI Grok 4.20 (reasoning)53LLMs
DeepSeek V3.258LLMs
Cohere Command A Translate58LLMs
NiuTrans60MT engines
Mistral Large 366LLMs
OpenAI GPT-5.6 Sol69LLMs
OpenAI GPT-5.573LLMs
Anthropic Claude Sonnet 575LLMs
AI21 Jamba Large 1.775LLMs
Anthropic Claude Fable 579LLMs
AppTek NMT80MT engines
Anthropic Claude Sonnet 4.685LLMs
Google Gemini 3.1 Pro Preview97LLMs
Anthropic Claude Opus 4.899LLMs
Alibaba Qwen3-235B-A22B157Open-weight LLMs
Google Gemma 4194Open-weight LLMs
z.ai GLM-5206LLMs
MoonshotAI Kimi K2.5250LLMs

Real-time models are those the shortlist counts as real-time, under 2 seconds per text across the languages we timed. The chart shows the time of each model into Macedonian.

Real-time

Eight of the nine real-time models in Macedonian are MT engines or LLM-based MT engines. Translations from the three on the shortlist, Microsoft Translator, DeepL Next Gen and Translated Lara 2, have 57.7, 43.2 and 24.7 weighted errors per 1,000 words. None of them meet general translation quality. The solutions that meet every Standard requirement are all on LLMs and take 32 to 97 seconds per 1,000 words, and the one that meets every Advanced requirement takes 64.

How we measure it: 12-01 · 12-02 · 12-09

Tone of voice and gender-neutral English to Macedonian translation

Tone of voice
Formal register, with Вие
Gender-neutrality strategy
neutral rephrasing + plurals/impersonals
Options in the language
generic masculine, feminine pairing or slash forms, neutral rephrasing
Default gender
None where the source leaves the gender open
The two best solutionsSolutionSentences about people that are gender-neutralMeets every Advanced requirementCost per million charactersSeconds per 1,000 words
OpenAI GPT-5.6 Terrathree-step, with the gender-neutrality agent99.1%✗$976160
OpenAI GPT-5.5one-step, with the gender-neutrality agent100%✓$12.5764
Trade-off

For tone of voice we check the formal register, with Вие as the form of address. For gender neutrality we check neutral rephrasing with plurals and impersonal forms, and no default gender where the source leaves it open. With the gender-neutrality agent, the three-step solution (translation + LQA + repair) on OpenAI GPT-5.6 Terra has 2.6 weighted errors per 1,000 words, and the one-step solution on OpenAI GPT-5.5 has 12.2. The first costs $976 per million characters and takes 160 seconds per 1,000 words, against $12.57 and 64.

How we measure it: 02-05 · 02-06 · 08-11 · 09-18

How confident are we in measuring English to Macedonian translation quality?

Confidence

Neither pair of automatic metrics is borderline in Macedonian, so the metric ranks behind the shortlist carry weight. The three LQA judges agree on a median of 95.5% of the sentences across the 11 shortlisted solutions.

How we measure it: 00-07 · 07-05 · 09-02

Automate translation to meet your requirements

See Intento Language Hub in action.

Book a demo
Leave feedback