Carbon Offsets vs Carbon Removals: Which Should You Buy in 2026?

The short answer: Carbon offsets fund activities that prevent emissions that would otherwise have occurred (avoidance). Carbon removals physically extract carbon dioxide already in the atmosphere. Serious net-zero commitments in 2026 require both — but the right ratio depends on what you are trying to achieve, who is asking you to prove it, and your budget per ton.

If you are new to this distinction, it sounds like semantics. It is not. Since the SBTi Net-Zero Standard v2 (January 2026) and the Oxford Offsetting Principles revision, the line between the two categories has become the single biggest factor determining whether a company can credibly claim net-zero alignment. For individuals, it determines whether your €10 donation does something permanent or something temporary.

The real difference between avoidance and removal

Carbon avoidance (what most "offsets" are)

An avoidance credit represents emissions that would have been released but were prevented. Classic examples:

• REDD+ forest protection — paying to keep a forest standing that would otherwise have been logged or converted to palm oil. One ton of CO₂ locked in wood that would have ended up in the atmosphere.
• Landfill gas capture — methane emitted by decomposing waste is combusted or turned into electricity instead of escaping. Methane is 28× worse than CO₂ over 100 years, so the leverage is huge.
• Renewable energy in grids still burning coal — a solar plant that displaces coal-fired generation. Historically the dominant credit type; increasingly excluded by high-integrity buyers because additionality is questionable in markets where solar is already cost-competitive.
• Efficient cookstoves — replacing open-fire cooking with enclosed stoves that burn 40–60% less fuelwood, which avoids both emissions and deforestation.

Carbon removal

A removal credit represents CO₂ that has been taken out of the atmosphere and stored somewhere it will not return. Examples:

• Afforestation and reforestation — planting new trees or restoring previously-deforested land. Low cost, nature-based, medium permanence (decades to ~100 years with proper management).
• Blue carbon — restoring mangroves, seagrass, and salt marshes. These coastal ecosystems store carbon 10× faster per hectare than tropical forest.
• Biochar — pyrolyzing agricultural residue into stable carbon that is buried in soil. Thousand-year permanence.
• Enhanced rock weathering — spreading crushed basalt on farmland, where it chemically binds atmospheric CO₂ over years to decades.
• Direct air capture with geological storage (DAC+S) — industrial-scale fans and chemical solvents pulling CO₂ from ambient air, then injecting it into deep saline aquifers. Permanence of tens of thousands of years. The gold standard — and the most expensive.

At a glance

	Carbon avoidance (offsets)	Carbon removal
What it does	Prevents future emissions	Extracts past emissions
Permanence	1–100 years (varies wildly)	100–10,000+ years
Typical cost per ton CO₂e	€2–30	€30–600
Volume available today	Billions of tons	~0.1% of avoidance volume
Best for	Scaling climate finance quickly	Residual emissions, durable claims
Oxford Principles category	Types 1 and 2	Types 3, 4, and 5

Why the distinction became critical in 2026

Three 2026 developments made this the year the offset-vs-removal debate matured:

1. SBTi Net-Zero Standard v2 — for the first time, companies claiming net-zero under SBTi must neutralize their residual emissions (the portion they cannot eliminate through operational reductions) using durable removals only. Avoidance credits are still permitted for interim "beyond value chain mitigation" claims, but cannot count toward the net-zero line itself.
2. EU Green Claims Directive enforcement — as of Q1 2026, any EU-marketed product or company claim involving the word "neutral" must disclose the mitigation type behind it. Avoidance-only claims must explicitly state "avoidance" and disclose permanence risk.
3. ICVCM Core Carbon Principles expanded coverage — the Integrity Council for the Voluntary Carbon Market tagged a first wave of methodologies as CCP-approved, creating the first widely-accepted quality floor. Methodologies that rely on counterfactual baselines (most avoidance) face a higher verification bar than direct measurement (most removals).

The Oxford Principles framework

The most useful framework for navigating this space is the Oxford Offsetting Principles, which classify credits into five types arranged on a ladder of durability:

• Type 1 — Emission reductions with short-lived storage: efficient cookstoves, landfill gas capture. High leverage, low permanence.
• Type 2 — Emission reductions with long-lived storage: REDD+ forest protection. Medium permanence, large volume.
• Type 3 — Removal with short-lived storage: afforestation, reforestation. Nature-based removal, decades-scale permanence.
• Type 4 — Removal with long-lived storage: blue carbon (mangroves), biochar. Centuries-scale permanence.
• Type 5 — Engineered removal with geological storage: DAC+S, enhanced weathering with measurement. Geologic-scale permanence.

Oxford's guidance: portfolios should shift up the ladder over time, increasing the share of Type 4 and 5 credits year after year. Many 2020 net-zero pledges were 100% Type 1 and 2. The 2026 norm for credible portfolios is 60–70% avoidance and 30–40% removal, trending toward 100% Type 4/5 removal by 2050.

We publish our full project mix — categorized by Oxford Type — in our verified impact portfolio.

When to buy which

For individuals

If you are offsetting a flight, a car commute, or annual personal emissions, the cost-per-ton reality matters. A one-way transatlantic flight generates about 1.5 tons of CO₂. At €5/ton avoidance credits, that is €7.50. At €400/ton DAC, that is €600. For most personal offsetting budgets, a realistic mix is 80% high-integrity avoidance (Oxford Types 1 and 2) plus 20% nature-based removal (Type 3). This is what most individual subscription services, including ours, default to.

For small and medium companies

Below about 10,000 tons per year, your scope 1–3 inventory is usually dominated by energy, travel, and supply chain — most of which cannot be reduced to zero in the short term. A credible 2026 portfolio for SMEs looks like: 50% Type 2 avoidance (REDD+), 30% Type 3 nature-based removal, 20% Type 4/5 durable removal. The Type 4/5 slice is expensive per ton but is what auditors look for when verifying your "neutral" claim.

For large corporates with SBTi commitments

SBTi v2 makes this almost prescriptive. Interim emissions reductions (pre-net-zero) can use any high-integrity credits for beyond-value-chain claims. But the final neutralization of residuals at target year must be 100% durable removals (Type 4 and 5). Most Fortune 500 climate teams are now signing long-term offtake agreements with DAC, enhanced weathering, and biochar providers to lock in supply before prices rise.

The price reality

Current 2026 retail prices for verified credits:

• Landfill gas (Type 1): €2–5 per ton
• REDD+ forest protection (Type 2): €5–15 per ton
• Reforestation (Type 3): €10–30 per ton
• Blue carbon mangroves (Type 4): €20–40 per ton
• Biochar (Type 4): €120–180 per ton
• Enhanced weathering (Type 5): €200–350 per ton
• DAC with geological storage (Type 5): €400–600 per ton

Our own platform sources from projects across every type. Prices are transparent and start at €0.04 per kg CO₂e (€40 per ton) for a high-integrity mixed portfolio that is weighted toward avoidance today and progressively re-weighted toward removal each year.

How to verify quality before you buy

Whatever category you pick, three checks separate a real credit from a greenwashing risk:

1. Standard: is the project verified under Verra VCS, Gold Standard, Puro.earth, Isometric, or CAR? If a seller cannot name the registry, walk away.
2. Vintage: when was the credit issued? Post-2020 vintages reflect tighter methodologies. Pre-2015 vintages, especially in renewable energy, are often of questionable additionality.
3. Retirement proof: can you verify the credit serial number was retired against your purchase in the public registry? Every legitimate retailer must provide this.

We publish all three data points for every ton we sell — read the full methodology for how we source, verify, and retire credits on your behalf.

Our recommended 2026 portfolio mix

Based on the three forces above (SBTi v2, EU enforcement, ICVCM integrity floor) and what Coffset's own portfolio research department published in March 2026, we recommend the following baseline allocation for most buyers in 2026:

• 40% Oxford Type 2 (REDD+ and nature-based avoidance) — scale, integrity, and measurable co-benefits for biodiversity and local communities.
• 25% Oxford Type 3 (reforestation and afforestation) — nature-based removal, visible co-benefits, moderate permanence.
• 20% Oxford Type 4 (blue carbon and biochar) — durable removal in the €20–180 range, the sweet spot of cost and permanence.
• 15% Oxford Type 5 (DAC, enhanced weathering) — the premium slice that makes your claim audit-proof.

Tilt the Type 5 share higher if you have SBTi commitments, lower if you are a budget-constrained individual. Avoid the pre-2026 convention of 100% Type 1 and 2 — it is increasingly unsupportable as a "neutral" claim in any regulated market.

Frequently asked questions

Are carbon offsets greenwashing?

Some are. The ones that claim additionality without proving it, the ones that double-count, the ones that use inflated baselines — those deserve the criticism. But a verified Type 2 REDD+ credit with a post-2020 vintage, retired against your purchase in a public registry is real mitigation. The answer is not to abandon offsetting; it is to buy the right credits and verify them.

Do I need to reduce emissions first?

Yes. The hierarchy is always avoid → reduce → offset/remove. Coffset explicitly positions offsetting as the third step. Use our carbon footprint calculator to identify reductions first; buy credits only for the residual you cannot reduce this year.

Will carbon removal get cheaper?

Yes, but not as fast as some think. DAC costs have dropped from ~€1,000/ton in 2020 to ~€450/ton today. IEA forecasts suggest ~€200/ton by 2035 and ~€100/ton by 2050. Enhanced weathering is on a faster curve. Biochar is already near its floor. For 2026 budgeting, plan for 10–15% annual cost decline, not 50%.

What about permanence risk for forests?

Real risk. A forest that burns releases its stored carbon. Modern REDD+ methodologies mitigate this with buffer pools (typically 10–20% of credits held back as insurance), monitoring, and legal durability (30–100 year project agreements with local governments and communities). No forest credit should be treated as permanent, but properly buffered credits have weathered multiple real fire events without claim erosion.

Can I offset with removals only?

Yes, but the math is brutal. Offsetting a 10-ton annual footprint at €400/ton DAC is €4,000 per person per year. For scale, most experts recommend the mixed portfolio above — heavy on high-integrity avoidance today, migrating progressively to removal over the coming decade as removal supply scales and costs drop.

What is Coffset's own mix?

Our default subscription portfolio for 2026 is 60% Type 2, 20% Type 3, 15% Type 4, 5% Type 5, at an average €40 per ton retail. Enterprise portfolios are custom-built toward SBTi-compliant residual neutralization. Full project-by-project breakdown is in our impact portfolio.

Bottom line

In 2026, the "offsets vs removals" question is settled in theory: you need both, with the ratio shifting toward removal over time. The hard part is execution — picking verified credits, getting the mix right for your goals, and avoiding the lower tier of the market. Start with our verified portfolio, read the methodology, and if you are ready, offset now from €0.04 per kg.